![0
1
2S Intel® Xeon® processor 2699 v4
Intel® Xeon® Platinum 8168 processor
39
PexipInfinity2.6*
Application:
Pexip Infinity is video communications software that organizations can
deploy across their IT infrastructure, in a public or private cloud or any
hybrid combination – so that every individual can have their own personal
high-quality video, audio and mobile meeting experience.
Potential Customer Benefits:
Higher number of 1080HD video streams that Intel® Xeon® 8168 Platinum
can handle (2x61 compared to 2x45 on E5-2699 v4, VP9) enables more
streams per server which keeps TCO down.
Performance Factors:
Manual and compiler Intel® AVX-512 enabling.
“The VP9 video codec with superblocks of 64 x 64 pixels is particularly well
suited to the Intel AVX-512 instruction set since a 512 bit registers can hold
exactly 64 pixels. At similar visual quality, our VP9 implementation is roughly
3 times faster than the LibVPX implementation that is not yet adopted to
AVX512.”
- Lars Petter Endresen, PhD, Principal Performance Engineer, Pexip
“The Intel Xeon Scalable processors with Intel AVX-512 provides up to
almost 50% performance increase for VP9 encoding and decoding
[compared to the previous generation Broadwell CPUs]. This translates to a
dramatic reduction in TCO for our customers whether they deploy in an IaaS
cloud or on-premises, and will enable very high quality enterprise video
communication at scale.” – Håkon Dahle, CTO, Pexip
ATAGLANCE
Hardware:
Intel® Xeon® Platinum 8168
processor
Platform Features:
Intel® Advanced Vector
Extensions 512 (Intel® AVX-
512)
Software Tools/Libraries:
Intel® Parallel Studio XE
2018
Up to
35%
faster
Pexip Infinity* increased performance with the
Intel® Xeon® Platinum 8168 processor1
1 - Testing conducted on Pexip* software comparing Intel® Xeon® Platinum 8168 processor to 2S Intel® Xeon® Processor E5-2699 v4. Testing by Pexip. For complete testing configuration details, see the Configurations Details section, slide 135.
MORE info
https://www.webmproject.or
g/code/vp9/
NormalizedPerformance
www.pexip.com
Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems,
components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated
purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others
Intel® Xeon® scalable Processor Family Intel® AVX-512
Workload: Pexip’s Infinity 2.6 video conferencing handling
maximum number of 1080HD video streams (withVP9 codec )
Video processingcomms](https://image.slidesharecdn.com/sklpurley075eamgprelimproofpointspublic-170619170758/85/Intel-Xeon-Scalable-Processors-Enabled-Applications-Marketing-Guide-39-320.jpg)
![146
CST EM Studio, time domain solver standard benchmark.
Baseline: 2S Intel® Xeon® processor E5-2697 v3, 2.6GHz, HT on, 14 cores, turbo on, 64GB DDR4, CentOS 6.4
Generation1: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, HT on, 18 cores, turbo on, BIOS GRRFSDP1.86B.0275.R01.1603300531, 64GB total memory, CentOS Linux* 6.8 kernel 2.6.32.
Generation2: 2S Intel® Xeon® Gold 6148 processor, 2.4GHz, HT on, 20 cores, turbo on, BIOS SE5C620.86B.01.00.0412.020920172159, 192GB total memory, CentOS Linux* 7.3.1611 kernel 3.10.0
SIMULIA Abaqus 2017x standard benchmarks tested by Intel, March 2017
BASELINE: 2S Intel® Xeon® processor E5-2698 v3, 2.3GHz, 16 cores, turbo and HT on, BIOS SE5C610.86B.01.01.5008.081020151115, 128GB total memory, 8 memory channels / 8x16GB / 2133 MT/s / DDR4,
Red Hat Enterprise Linux* 7.3, kernel 3.10.0-229.20.1.el6.x86_64.knl2, Intel® SSD S3700 Series.
NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo and HT on, BIOS GRRFSDP1.86B0271.R00.1510301446, 128GB total memory, 8 memory channels / 8x16GB / 2400 MT/s / DDR4,
Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327.el7.x86_64, 2x Intel® SSD S3500 Series RAID0 array.
NEW: 2S Intel® Xeon® Gold processor 6148, 2.4GHz, 20 cores, turbo boost tech on, turbo energy efficient off, HT on, BIOS SE5C620.86B.01.00.0412.020920172159, 192GB total memory, 12 memory
channels / 12x16GB / 2666 MT/s / DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327.el7.x86_64, 2x Intel® SSD S3500 Series RAID0 array.
Note*: MKL shipped with 2017x Abaqus release is swapped with MKL 2017u1, and added environment variable: MKL_ENABLE_INSTRUCTIONS=AVX512 [Not an officially supported software kit]
SIMULIA Abaqus 2017x Explicit benchmarks tested by Intel, April 2017
BASELINE: 2S Intel® Xeon® processor E5-2697 v3, 2.6GHz, 14 cores, turbo and HT on, 128GB total memory, 8 memory channels / 8x16GB / 2133 MT/s / DDR4, Red Hat Enterprise Linux* 7.3, kernel 3.10.0-
514.6.2.0.1.el7.x86_64.knl1, Intel® SSD S3700 Series.
NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo and HT on, BIOS GRRFSDP1.86B0271.R00.1510301446, 128GB total memory, 8 memory channels / 8x16GB / 2400 MT/s / DDR4,
Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327.el7.x86_64, 2x Intel® SSD S3500 Series RAID0 array.
NEW: 2S Intel® Xeon® Gold processor 6148, 2.4GHz, 20 cores, turbo boost tech on, turbo energy efficient off, HT on, BIOS SE5C620.86B.01.00.0412.020920172159, 192GB total memory, 12 memory
channels / 12x16GB / 2666 MT/s / DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327.el7.x86_64, 2x Intel® SSD S3500 Series RAID0 array.
LSTC LS-DYNA
Baseline: 2S Intel® Xeon® processor E5-2697 v3, 2.3GHz, 14 cores, turbo and HT on, BIOS SE5C610.86B.01.01.5008.081020151115, 128GB total memory, 8 memory channels / 8x16GB / 2133 MT/s / DDR4,
Red Hat Enterprise Linux* 7.3, kernel 3.10.0-229.20.1.el6.x86_64
Next Gen: Dual Socket Intel® Xeon® processor E5-2697 v4 2.3 GHz, Turbo mode ON , 18 Cores/Socket, 36 Cores (HT off), DDR4 128GB, 2400 MHz, Wildcat Pass Platform. Disk: 800GB Intel SSD Kernel: 3.10.0-
229.20.1.el6.x86_64
New: Dual Socket Intel® Xeon® Gold 6148 processor 1.8 GHz, Turbo mode ON , 28 Cores/Socket, 56 Cores (HT off), DDR4 128GB, 2667 MHz, Wolf Pass Platform. Disk: 800GB Intel SSD Kernel:3.10.0-
327.e17.x86-64
INTES PERMAS: Baseline: 2S Intel® Xeon® processor E5-2697 v3, 2.6GHz, 14 cores, turbo on, HT off, NUMA on, BIOS 46.R00, 256 GB total memory (8x 32GB w/ 2400 MT/s, DDR4 LRDIMM), 4x Intel® SSD DC
P3600 1.6 TB in RAID 0 (stripe size 64k). CentOS* Linux* 6.6, kernel 4.1.2. Intel® Composer 2015.2.164. INTES PERMAS V15.10.10.
Next generation: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo on, HT off, NUMA on, BIOS 338.R00, 256 GB total memory (8x 32GB w/ 2400 MT/s, DDR4 LRDIMM), 4x Intel® SSD DC P3600 2
TB in RAID 0 (stripe size 64k). CentOS Linux* release 7.2, kernel 3.10.0-327.13.1.el7.x86_64. Intel® Composer 2015.5.223. INTES PERMAS V16.00.
New: 2S Intel® Xeon® Gold processor 6148, 2.4 GHz, 20 cores, turbo on, HT off, NUMA on, BIOS SE5C620.86B.01.00.0412.020920172159, 384 GB total memory (12x 32GB w/ 2400 MT/s, DDR4 LRDIMM), 3x
Intel® SSD DC P3600 2 TB in RAID 0 (stripe size 64k), CentOS* Linux* release 7.3, kernel 3.10.0-514.10.2.el7.x86_64. Intel® Composer 2015.7.235. INTES PERMAS V16.00.
INTES PERMAS (2 socket): E5-2680: 2S Intel® Xeon® processor E5-2680, 2.7 GHz, 8 cores, turbo on, HT off, NUMA on, 160 GB total memory (8x 16 GB & 8x 4 GB w/ 1333 MT/s, DDR3), 2x SSD OCZ Vertex 2 in
RAID 0 (stripe size 64k) and 6x Seagate ST3300657SS (stripe size 256k). CentOS* Linux* 7.1, kernel 3.10.0-229.20.1, Intel® Composer 2015.2.164. INTES PERMAS V15.10.12.
E5-2697v2: 2S Intel® Xeon® processor E5-2697 v2, 2.7 GHz, 12 cores, turbo on, HT off, NUMA on, BIOS 28D690, 128 GB total memory (8x 16 GB w/ 1600 MT/s, DDR3), 4x SSD Samsung 840pro in RAID0
(stripe size 64k), CentOS* Linux* 6.6, kernel 2.6.32-573.7.1, Intel® Composer 2011_sp1.13.367. INTES PERMAS V15.10.12.
E5-2697v3: 2S Intel® Xeon® processor E5-2697 v3, 2.6GHz, 14 cores, turbo on, HT off, NUMA on, BIOS 46.R00, 256 GB total memory (8x 32GB w/ 2400 MT/s, DDR4 LRDIMM), 4x Intel® SSD DC P3600 2 TB in
RAID 0 (stripe size 64k). CentOS* Linux* 6.6, kernel 4.1.2. Intel® Composer 2015.2.164. INTES PERMAS V15.10.10.
E5-2697v4: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo on, HT off, NUMA on, BIOS 338.R00, 256 GB total memory (8x 32GB w/ 2400 MT/s, DDR4 LRDIMM), 4x Intel® SSD DC P3600 2 TB in
RAID 0 (stripe size 64k). CentOS Linux* release 7.2, kernel 3.10.0-327.13.1.el7.x86_64. Intel® Composer 2015.5.223. INTES PERMAS V16.00.
6148: 2S Intel® Xeon® Gold processor 6148, 2.4 GHz, 20 cores, turbo on, HT off, NUMA on, BIOS SE5C620.86B.01.00.0412.020920172159, 384 GB total memory (12x 32GB w/ 2400 MT/s, DDR4 LRDIMM), 3x
Intel® SSD DC P3600 1.6 TB in RAID 0 (stripe size 64k), CentOS* Linux* release 7.3, kernel 3.10.0-514.10.2.el7.x86_64. Intel® Composer 2015.7.235. INTES PERMAS V16.00.](https://image.slidesharecdn.com/sklpurley075eamgprelimproofpointspublic-170619170758/85/Intel-Xeon-Scalable-Processors-Enabled-Applications-Marketing-Guide-146-320.jpg)
![150
VASP CONFIGURATION: Beta VASP, a release candidate for v6.0. Developer branch provided as “Package” included with download: https://github.com/vasp-dev/vasp-knl. AVX512: Intel® Compiler 17.0.1.132,
Intel® MPI 2017u1, ELPA 2016.05.004. Optimization Flags: “-O3 -xCORE-AVX512“. AVX2: Intel® Compiler 17.0.1.132, Intel® MPI 2017u1, ELPA 2016.05.004. Optimization Flags: “-O3 -xCORE-AVX2“.
BASELINE CONFIGURATION: 2S Intel® Xeon® processor E5-2699 v3 2.3 GHz, 18 Cores/Socket, 36 Cores, 72 Threads, HT on, turbo off, 128GB total memory, 2133 MT/s / DDR4 RDIMM, Red Hat Enterprise
Linux* 7.0 kernel.
NEXT GEN CONFIGURATION: 2S Intel® Xeon® processor E5-2697 v4 2.3 GHz , 18 Cores/Socket, 36 Cores, 72 Threads, HT on, turbo off, BIOS 86B0271.R00, 128GB total memory, 2400 MT/s DDR4 RDIMM, Red
Hat Enterprise Linux* 7.2 kernel 3.10.0-327.
NEW CONFIGURATION: 2S Intel® Xeon® Gold 6148 processor 2.4 GHz , 20 Cores/Socket, 40 Cores, 80 Threads, HT on, turbo off, BIOS 86B.01.00.0412, 192GB total memory, 2666 MT/s / DDR4 RDIMM, Red
Hat Enterprise Linux* 7.2 kernel 3.10.0-327.
Amber: Version 16 with all patches applied at December, 2016. Workloads: PME Cellulose NVE(408K atoms), PME stmv(1M atoms), GB Nucleosome (25K), GB Rubisco (75K). No cut-off used for GB workloads.
Compiled with -mic2_spdp –intelmpi - openmp, –DMIC2 defined. DMIC2 enable optimization for AVX512 vectorization, SPDP mixed precision, OpenMP optimization.Tests performed on March 2017.
BASELINE: Executed with 36 MPI, 2 OpenMP. 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 36 cores, turbo and HT on, BIOS 86B0271.R00, 8x16GB 2400MHz DDR4, Red Hat Enterprise Linux* 7.2 kernel
3.10.0-327.
NEW: Executed with 40 MPI and 2 OpenMP. 2S Intel® Xeon® Gold 6148 processor, 2.4GHz, 40 cores, turbo on, HT on, BIOS 86B.01.00.0412.R00, 12x16GB 2666MHz DDR, Red Hat Enterprise Linux* 7.2 kernel
3.10.0-327.
NAMD: Version 2.12 Dec2016. Workloads: apoa1(92K atoms), stmv(1M atoms). Compiled with –DNAMD_KNL* define. Tests performed on March 2017.
BASELINE: Executed with 72 charm threads. 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 36 cores, turbo and HT on, BIOS 86B0271.R00, 8x16GB 2400MHz DDR4, Red Hat Enterprise Linux* 7.2 kernel
3.10.0-327. Compiler option “-xCORE-AVX2”.
NEW: Executed with 40 charm threads. 2S Intel® Xeon® Gold 6148 processor, 2.4GHz, 40 cores, turbo on, HT on, BIOS 86B.01.00.0412.R00, 12x16GB 2666MHz DDR, Red Hat Enterprise Linux* 7.2 kernel
3.10.0-327. Compiler option “-xCORE-AVX512”.
QMCPACK CONFIGURATION: Developer branch provided as “rebuild-dirac.soa-fix-distAA” downloaded from: https://xgitlab.cels.anl.gov/QMCPACK/qmcpack-dev.git, Date: 04/05/2017. Intel® parallel studio
17.2.174, Optimization Flags: “-o3 -ip -restrict -unroll -g -debug inline-debug-info -openmp -std=c++11”.
BASELINE CONFIGURATION: 2S Intel® Xeon® processor E5-2697 v4 2.3 GHz , 18 Cores/Socket, 36 Cores, 72 Threads, HT on, turbo off, BIOS 86B0271.R00, 128GB total memory, 2400 MT/s DDR4 RDIMM, Red
Hat Enterprise Linux* 7.2 kernel 3.10.0-327. (-xCORE-AVX2), run with 2088 walkers, 2 MPI with 36 threads each.
NEW CONFIGURATION: 2S Intel® Xeon® processor Gold 6148 processor 2.4 GHz , 20 Cores/Socket, 40 Cores, 80 Threads, HT on, turbo on, BIOS 86B.01.00.0412, 192GB total memory, 2666 MT/s DDR4
RDIMM, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. (-xCORE-AVX512), run with 2080 walkers, 2 MPI with 40 threads each.
Portable code transformation to increase SIMD efficiency and reduce memory footprint [1,2] The efforts result in 3.1x-4.3x improvements on 2S Intel® Xeon® processor Gold 6148 and 2.6x-6.3x improvements
on 2S Intel® Xeon® processor E5-2697 v4 processor from baseline code. These speedups are wrt. QMPCACK release 3.0.0 on the same hardware with QMC_MIXED_PRECISION=0 flag.
Monte Carlo workload. OS: Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. Testing by Intel March 2017.
BASELINE: 2S Intel® Xeon® processor E5-2697 v3, 2.6GHz, 28 cores, turbo and HT on, BIOS 86B.0036.R05, 64GB total memory, 8x8GB 2133 MHz DDR4, Fedora release 20 kernel 3.15.10-200 .
NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4 , 2.3GHz, 36 cores, turbo and HT on, BIOS 86B0271.R00, 128GB total memory, 8 x16GB 2400 MHz DDR4 RDIMM, 1 x 1TB SATA, Red Hat Enterprise Linux* 7.2
kernel 3.10.0-327.
NEW: 2S Intel® Xeon® Gold 6148 processor @ 2.4GHz, H0QS, 40 cores 150W. QMS1, turbo and HT on, BIOS SE5C620.86B.01.00.0412.020920172159, 192GB total memory, 12 x 16 GB 2666 MHz DDR4
RDIMM, 1 x 800GB Intel® SSD SC2BA80, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327
Black-Scholes workload. OS: Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. Testing by Intel March 2017.
BASELINE: 2S Intel® Xeon® processor E5-2697 v3, 2.6GHz, 28 cores, turbo and HT on, BIOS 86B.0036.R05, 64GB total memory, 8x8GB 2133 MHz DDR4, Fedora release 20 kernel 3.15.10-200 .
NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4 , 2.3GHz, 36 cores, turbo and HT on, BIOS 86B0271.R00, 128GB total memory, 8 x16GB 2400 MHz DDR4 RDIMM, 1 x 1TB SATA, Red Hat Enterprise Linux* 7.2
kernel 3.10.0-327.
NEW: 2S Intel® Xeon® Gold 6148 processor @ 2.4GHz, H0QS, 40 cores 150W. QMS1, turbo and HT on, BIOS SE5C620.86B.01.00.0412.020920172159, 192GB total memory, 12 x 16 GB 2666 MHz DDR4
RDIMM, 1 x 800GB Intel® SSD SC2BA80, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327](https://image.slidesharecdn.com/sklpurley075eamgprelimproofpointspublic-170619170758/85/Intel-Xeon-Scalable-Processors-Enabled-Applications-Marketing-Guide-150-320.jpg)
Uploaded byIntel IT Center
Intel® Xeon® Scalable Processors Enabled Applications Marketing Guide
The document provides an overview of the Intel® Xeon® Scalable Processors, highlighting their role in enhancing performance across various software ecosystems, including cloud, communications, enterprise applications, and high-performance computing. It details specific performance increases compared to previous generations and showcases software proof points, demonstrating significant gains in processing power and efficiency. Additionally, it emphasizes the processors' capabilities in data center innovation and infrastructure agility, with various testimonials and use cases from industry partners.
More Related Content
Similar to Intel® Xeon® Scalable Processors Enabled Applications Marketing Guide
Intel® Xeon® Scalable Processors Enabled Applications Marketing Guide
- 1. July 2017 –Intel Software Services Group Intel® Xeon® ScalableProcessors EnabledApplicationsmarketingguide
- 2. 2 Intel®Xeon®scalableprocessors EAMGTableofContents SoftwareEcosystem Summary/Highlights Platform Value Software Proof Points: Cloud Comms Enterprise High Performance Computing 2x2s Helpful Links Configurations & Disclaimers Slide 3 Slide 5 Slide 6 Slide 9 Slide 10 Slide 25 Slide 39 Slide 68 Slide 135 Slide 138 Slide 139
- 3. 3 BroadSoftwareEcosystemIntel®xeon®ScalablePlatformsupport System R&D Centerof NARI Technology co., LTD *Other names and brands may be claimed as the property of others.
- 4. SoftwarePartnersEnablingonIntel® Xeon® scalableprocessors 4 Comms 6Wind,AsiaInfo, Beamr, eBrisk, Ericsson, Keepixo, MulticoreWare, Pexip, Shenzhen TYDIC, Sandvine Cloud/ Virtualization Baidu, Citrix, Huawei, LeTV, Kingsoft, Neusoft, Qihoo, Tencent, TongTech, Virtuozzo, VMware, Zizo Enterprise Altibase, Aerospike. Alibaba, Compuverde, Exasol, IBM, Intersystems, Innovatrics, Kingbase, Kingdom Technology, NariTech, NRI, Oracle, Qlik, QuikFynd, SAP, SAS, Software AG, Sunjesoft, Yonyou, Zebware HPC DIGITAL CONTENT CREATION Intel Embree, Pixar ENERGY Mines ParisTech FINANCIAL SERVICES IHS Markit, Itivity LIFE SCIENCES Gaussian, Sandia Corp, Pacific NW Nat’l Lab MANUFACTURING Altair, ANSYS, Convergent Science, CST, Dassault Systemes, ESI Group, Free Field Technologies, Intes Software, KAIST/KISTI, LSTC, MSC Software, S & I Engineering, Siemens, Xidian University WEATHER First Institure of Oceanography OSV Canonical, Microsoft, Red Hat, SUSE *Other names and brands may be claimed as the property of others.
- 5. 5 RE-architectingthedatacenterwithIntel®Xeon®scalableProcessors HPC Gen-to-Gen Performance Increase – Slide69 Cloud Gen-to-Gen Performance Increase – Slide 17 Communications Services Gen-to-Gen Performance Increase – Slide 28 Enterprise Gen-to-Gen Performance Increase – Slide 42 More Information: www.intel.com/XeonSoftwareSolutions 2.14XUp to 1.9XUp to 1.74XUp to 1.61XUp to Performance.Security.agility Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others 1 1 1 1 1 – As demonstrated by proof points in this presentation
- 6.
- 7. Intel®Xeon®ScalableplatformThe foundation ofData Center Innovation: Agile & Trusted Infrastructure delivers1.65xaverageperformanceboostoverpriorGeneration1 7 Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more information go to http://www.intel.com/performance Intel does not control or audit the design or implementation of third party benchmark data or Web sites referenced in this document. Intel encourages all of its customers to visit the referenced Web sites or others where similar performance benchmark data are reported and confirm whether the referenced benchmark data are accurate and reflect performance of systems available for purchase. 1 - Up to 1.65x claim based on STREAM - triad: 1-Node, 2 x Intel® Xeon® Processor E5-2699 v4 on Grantley-EP (Wellsburg) with 256 GB Total Memory on Red Hat Enterprise Linux* 6.5 kernel 2.6.32-431 using Stream NTW avx2 measurements. Data Source: Request Number: 1709, Benchmark: STREAM - Triad, Score: 127.7 Higher is better vs. 1-Node, 2 x Intel® Xeon® Platinum 8180 Processor on Neon City with 384 GB Total Memory on Red Hat Enterprise Linux* 7.2-kernel 3.10.0-327 using STREAM AVX 512 Binaries. Data Source: Request Number: 2500, Benchmark: STREAM - Triad, Score: 199 Performance Pervasive through compute, storage, and network Agility Rapid service delivery Security Pervasive data security with no performance overhead
- 8. AGlimpseInsidethe Intel®Xeon®Scalableplatform 8 SSDs Intel® Optane™ SSD DCP4800X Complementary Intel® FPGA INTEGRATED OPTIONS Fabric Intel® Omni-Path Architecture Networking Intel® Ethernet Accelerators Intel® QuickAssist Intel® AVX-512 Workload optimized frameworks & telemetry (e.g. Caffe*, Intel® DAAL, Intel® MKL, DPDK, SNAP*, SPDK) Advancingvirtuallyeveryaspect: BrandNewcore, cache,on-dieinterconnects,memorycontroller&more performance Agilitysecurity Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Intel® Volume Management Device (Intel® VMD) Intel® Data Analytics Acceleration Library (Intel® DAAL) Intel® Math Kernel Library (Intel® MKL) Storage Performance Development Kit (SPDK) Data Plane Development Kit (DPDK) Intel® Resource Director Technology (Intel® RDT)
- 9. software 9 • CloudServiceProviders • CommunicationsServices •enterpriseapplications • highperformancecomputing
- 10.
- 11. 11 Intel®Xeon®ProcessorScalableFamily Performance Proof Points– Cloud Potential Customer Benefits: • Reduced latency for an improved user experience • Accelerated processing to support novel products • Enhanced responsiveness for improved service delivery cloud 1.74XUp to GENERATION-TO-GENERATION INTEL® AVX-512 1.34XUp to INTEL® OPTANE™ TECHNOLOGY INTEL® QUICKASSIST TECHNOLOGY 1.63XUp to 2XUp to VIEW 1 1 1 1 1 – As demonstrated by proof points in this presentation Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Average Gen-to-Gen processor application speedup: 1.75X See slide 17 See slide 20 See slide 18 See slide 15
- 12. Metric: geomean fromSPECjbb and WebBench throughputs 12 VirtuozzoSPECjbb+WebBench* Application: Virtuozzo is used by over 700 service providers, ISVs and enterprises to enable over 5 million virtual environments running mission-critical cloud workloads. The test was conducted using combination of SPECjbb and WebBench benchmarks, results being reported as geometric mean of their throughput rate. Potential Customer Benefits: Cloud Service Providers can perform services faster and more efficiently for a better customer experience. Performance Factors: By taking advantage of new micro-architecture and memory bandwidth from Intel® Xeon® Scalable processors, Virtuozzo was able to achieve significant performance gains with both its Container and Hypervisor solutions. The Intel® Xeon® Platinum 8168 processor delivered up to 2.2X1 better performance with Containers and up to 3.2X1 better performance for up to 32 pairs of Virtuozzo Hypervisor* instances. Up to 100% faster 1 - Testing conducted on Virtuozzo* software comparing Intel® Xeon® Platinum 8168 processor to Intel® Xeon® Processor E5-2680 v4. Testing by Virtuozzo. For complete testing configuration details, see the Configurations Details section, slide 135. www.virtuozzo.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable processorsVirtualization Virtuozzo Containers* increased performance up to 2.2X and Virtuozzo Hypervisor* up to 3.2X with the 2S Intel® Xeon® Platinum 8168 processor compared to the 2S Intel® Xeon® processor E5-2680 v4 Virtuozzo Containers Virtuozzo Hypervisor Cloud Up 3.2X faster Up 2.2X faster
- 13. 0 1 2 2S Intel® Xeon®processor E5-2699 V4 2S Intel® Xeon® Platinum 8180 Processor Up to 2.44X faster 13 HuaweiFusionSphere Application: Huawei FusionSphere is a backend virtualized cloud solution. FusionSphere uses virtualization technology to virtualize the backend server resources and provides agile and efficient HW infrastructure environment for enterprise solutions. Potential Customer Benefits: Support more VMs to parallelly finish business operations within given performance criteria. Accelerate processing with Intel® AVX-512 technology Performance Factors: Software optimizations for the Intel® Xeon® Platinum 8180 processor and Intel® AVX-512 delivered a 2.44X1 performance improvement compared to Intel® Xeon® E5-2699 v4 and Intel® AVX2. Huawei FusionSphere is fully optimized for Intel® AVX512 and Intel® AVX2. “The Intel® Xeon® Platinum 8180 processor is a great platform to support the virtualized infrastructure for our product, and it is one of the ideal deployment platforms for NC.” Chen Kun, VP of Huawei IT Cloud Computing Product Line “Intel® Xeon® Platinum 8180是我们应用的强大的虚拟底层平台，是我们 产品的理想部署平台之一。”, 陈锟, 华为IT云计算产品线副总裁 Huawei FusionSphere increased performance with the Intel® Xeon® Platinum 8180 processor with Intel® AVX-512 NormalizedPerformance www.huawei.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable processors Workload: HPC-like test-case simulates compute intensive applications, measuring Intel AVX-512 operations per second ATAGLANCE Hardware: Intel® Xeon® Platinum 8180 processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® Math Kernal Library Intel® MPI Intel® AVX-512 1 - Testing conducted on Fusionsphere* software comparing Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4 Testing by Huawei and Intel. For complete testing configuration details, see the Configurations Details section, slide 135. VirtualizationCloud
- 14. 0 1 2S Intel® Xeon®processor E5-2699 V4 2S Intel® Xeon® Platinum 8180 Processor 14 HuaweiFusionSphere* Application: Huawei FusionSphere is a backend virtualized cloud solution. FusionSphere uses virtualization technology to virtualize the backend server resources and provides agile and efficient HW infrastructure environment for enterprise solutions. Potential Customer Benefits: Support more VMs to parallelly finish business operations within given performance criteria. Performance Factors: More cores and high Required Frequency of the Intel® Xeon® Scalable processors platform were key to the overall performance gain of 1.62X1 compared to the previous processor generation. “The Intel® Xeon® Platinum 8180 processor is a great platform as virtualized infrastructure for our product, and it is one of the ideal deployment platforms for NC.” Chen Kun, VP of Huawei IT Cloud Computing Product Line “Intel® Xeon® Platinum 8180是我们应用的强大的虚拟底层平台，是我们产品的理想部署平台之一。”, 陈锟, 华为IT云计算产品线副总裁 Huawei FusionSphere increased performance with the Intel® Xeon® Platinum 8180 processor 1 - Testing conducted on Fusionsphere* software comparing Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4 Testing by Huawei and Intel. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedPerformance www.huawei.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable processors Workload: FusionSphere simulation of enterprise applications Up to 62% faster VirtualizationCloud
- 15. 0 1 2 Gen-to-Gen Intel® QuickAssist Technology 2SIntel® Xeon® processor E5-2699 v4 2S Intel® Xeon® Platinum 8180 Processor 2S Intel® Xeon® Platinum 8180 processor w/Intel® QAT 15 NeusoftSaCa®AgileCloud ManagementEnvironment(Aclome) Cloud Application: Neusoft SaCa Aclome is a visualized, automatic and agile cloud management environment enabling customers to deploy, monitor and manage cloud applications in a flexible, convenient manner. Potential Customer Benefits: Deliver more services in less time and facilitate a better user experience Accelerate compression/de-compression of user data with Integrated Intel® QuickAssist Technology (Intel® QAT) Performance Factors: More cores and high memory bandwidth with the Intel® Xeon® Platinum 8180 processor compared to the previous processor generation are key to the performance gain Integrated Intel® QAT speeds up compression/de-compression of user data. “We are excited to see the 50% performance gain on the new Intel® Xeon® Platinum 8180 processor compared to previous the previous generation, as well as the significant performance benefit from integrated Intel® QAT, which will definitely enable us to deliver better user experiences to our customers.” ZhiGang Huang, Vice General Manager of Neusoft Platform Software Division “跟之前的 Intel® Xeon® E5-2699 v4相比，基于 Intel® Xeon® Platinum 8180, 我 们非常高兴看到 SaCa Aclome 1.5倍的性能提升， 特别是Intel® QAT对压缩/解 压缩的大幅性能提升也十分令人满意。这将极大支持我们给最终用户提供更好 的用户体验”，黄志刚，东软基础软件事业部副总经理 Neusoft SaCa® Aclome* increased performance with the Intel® Xeon® Platinum 8180 processor with integrated Intel® QAT1 1 - Testing conducted on Neusoft* software comparing 2S Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4. Testing done by Intel. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedPerformance http://www.neusoft.com/ Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable processors Performance Metric: Throughput (Gen-to-Gen: transaction/second, QAT: Mbps: M-bit/second) Up to 2X ATAGLANCE Hardware: 2S Intel® Xeon® 8180 Platinum Processor Platform Features: Integrated Intel® Quick Assist Technology (Intel® QAT) Software Tools/Libraries: Intel® Math Kernel Library Intel® QAT Up to 1.5X
- 16. 0 1 4S Intel® Xeon®processor E7-8890 v4 4S Intel® Xeon® Platinum 8168 processor 16 ZizoAnalyticsPlatform* Analytics Application: Zizo is an In-Memory pattern database designed to run complex analytical workloads. Zizo can deliver cloud analytics and data science as a service through patented technology. Potential Customer Benefits: Faster response time in the front-end dashboard. Opportunity to use this extra time for speculative and predictive processing. Improved scalability and performance across large datasets and workloads. Performance Factors: By taking advantage of the new Intel® Xeon® Scalable processors microarchitecture , Zizo experienced a 92% gain1 compared to the previous processor generation. The gain came despite using an off the shelf Java Runtime. “As an in-memory database platform, delivering analytics at scale, we expect to deliver even better performance and scalability for all of our customers & partners on the Purley platform.” - Matthew Napleton, CMO at Zizo Up to 92% faster Zizo increased performance with the Intel® Xeon® Platinum 81681 1 - Testing conducted on Zizo* software comparing 4S Intel® Xeon® Platinum 8168 processor to 4S Intel® Xeon® Processor E7-8890 v4 Testing by Intel. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedPerformance www.zizo.co.uk Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Workload: Representative retail dashboard queries 4 socket Cloud
- 17. 0 1 2S Intel® Xeon®processor E5-2699 v4 2S Intel® Xeon® Platinum 8180 processor 17 BaiduSearchClick-Through-Rate Application: Baidu is the leading Chinese search engine, and is widely used in PRC. Click-Through-Rate (CTR) is one of the core modules of Baidu search. Potential Customer Benefits: Deliver more TPS to provide a better user experience Reduced latency in search engine Performance Factors: More cores and high memory bandwidth of the platform were key to the overall performance gain of 1.74X1 over the previous processor generation. “As one of the core modules of Baidu Search, the Click-Through Rate (CTR) sub-module gained a great performance improvement on Intel® Xeon® Platinum 8180 processor. This helps us a lot to improve our user experience.” Hou Jue, Baidu NF Architect “作为百度搜索的核心模块之一，Intel® Xeon® Platinum 8180帮助我们的CTR展示了强大的性能，必将大大提 升我们的用户体验”, 侯珏，百度基础架构部架构师． Baidu Search Click-Through-Rate (CTR)* increased performance with the Intel® Xeon® Platinum 8180 processor 1 - Testing conducted on Baidu* software comparing Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4 Testing by Intel. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedPerformance Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Workload: Baidu Search Click-Through-Rate transaction per hour www.baidu.com Up to 74% faster Intel® Xeon® Scalable processorsCloud
- 18. 0 1 2S Intel® Xeon®processor E5-2699 v4 2S Intel® Xeon® Platinum 8180 processor 18 Tencent*Videostitching media Application: Tencent video stitching application allows customers to create immersive video content for VR, 360 videos and other needs, based on certain algorithms. Potential Customer Benefits: Reduced latency in stitching videos together. Seamless, smoother and more immersive 360-degree video experiences. Performance Factors: More cores and high memory bandwidth from the Intel® Xeon® Scalable processor platform were key to the overall performance gain of 1.72X1 compared to the previous generation processor. Software optimizations leveraging Intel® AVX-512 delivered a 34%1 performance improvement compared to Intel® AVX2 instructions, which greatly enhanced the stitching function. “Glad to see the great performance of Intel® Xeon® Platinum 8180 processor, this will help us to deliver better user experience to our customers.” Weihua Jian, Manager, Tencent Video Processing Team “很高兴看到Intel® Xeon® Platinum 8180 processor 的强大性能，这将有助于为 我们的最终客户提供更好的客户体验”， 简伟华， 腾讯视频研发经理 Up to 72% faster Tencent* video stitching increased performance with the Intel® Xeon® Platinum 8180 processor 1 - Testing conducted on Tencent* software comparing Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4 Testing by Intel. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedPerformance Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others ATAGLANCE Hardware: 2S Intel® Xeon® 8180 Platinum Processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX- 512) Software Tools/Libraries: Intel® Math Kernel Library Intel® AVX-512 www.tencent.com Workload: Tencent video stitching Intel® Xeon® Scalable processorsCloud
- 19. 0 1 2S Intel® Xeon®processor E5-2699 v4 2S Intel® Xeon® Platinum 8180 w/Intel® AVX2 2S Intel® Xeon® Platinum 8180 w/Intel® AVX-512 19 TencentInGamePurchaseMachine LearningPlatform Application: Tencent InGame Purchase Machine Learning Platform predicts a user’s in-game purchase manners and provides purchase recommendations. Potential Customer Benefits: Ability to deliver a better user experience by providing more relevant purchase recommendations Performance Factors: More cores & high memory bandwidth of the platform were key to the overall performance gain of 1.57X over previous generation processor.1 Leveraging Intel® AVX-512 instructions delivered 26% better performance compared to Intel® AVX2.1 “Great to see 57% performance improvement with the Intel® Xeon® Platinum 8180 processor compared to the previous generation. This helps us to provide better, more relevant purchase recommendations to our customers.” Deng Dafu, Director, Operation Deputy, IEG “很高兴看到Intel® Xeon® Platinum 8180可以将我们产品的性能跟上一代平台相 比提升57%。这将帮助我们为客户提供更好的用户体验。”，邓大付, IEG互动娱 乐运营部数据挖掘应用中心总监 ATAGLANCE Hardware: 2S Intel® Xeon® 8180 Platinum Processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX- 512) Software Tools/Libraries: Intel® Math Kernel Library Tencent InGame Purchase Machine Learning Platform* increased performance with the Intel® Xeon® Platinum 8180 processor and Intel® AVX-512 1 - Testing conducted on Tencent* software comparing Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4 Testing by Intel. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedPerformance Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Workload: Tencent InGame purchase machine learning platform training time (seconds) Up to 57% faster Up to 24% faster Intel® AVX-512 www.tencent.com 26% w/Intel® AVX-512 machinelearning Intel® Xeon® Scalable processorsCloud
- 20. 1.30 1.55 1.63 0 1 Gen-to-Gen Intel® Optane™ Technology 2SIntel® Xeon® processor E5-2699 v4 2S Intel® Xeon® Platinum 8180 w/Intel® AVX2 2S Intel® Xeon® Platinum 8180 w/Intel® AVX-512 2S Intel® Xeon® Platinum 8180 w/Intel® P3700 2S Intel® Xeon® Platinum 8180 w/Intel® Optane™ 20 Kingsoftcloudimageprocessing&MySQL cloudservice Application: Kingsoft Cloud provides Image Cloud services, including image storage and processing, for its customers. MySQL Cloud is a new business for Kingsoft Cloud, which provides a simple, automated, integrated and enterprise ready cloud service, enabling organizations to increase business agility and reduce costs. Potential Customer Impact: Faster image processing provides ability to deliver a better user experience. Performance Factors: More cores with the Intel® Xeon® Platinum 8180 processor is key to the overall performance gain of 1.55X1 over the previous generation processor. Leveraging Intel® AVX-512 instructions delivered a 19% performance improvement compared to Intel® AVX2. High-speed, low-latency access to data with the Intel® Optane™ Solid State Drive DC P4800X dramatically improves I/O efficiency (63% improvement over the Intel® P3700 SSD)1 Kingsoft Cloud Image Processing* and MySQL Cloud Service* increased performance with Intel® Xeon® Platinum 8180 processor and Intel® Optane™ SSD 1 - Testing conducted on Kingsoft* software comparing Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4 Testing done by Intel. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedPerformance http://www.ksyun.com/ Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Workload: Frame per second (gen/gen) Intel Optane: Transaction/second Intel® AVX-512 ATAGLANCE Hardware: Intel® Xeon® 8180 Platinum Processor Intel® Optane™ Solid State Drive DC P4800X Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® Math Kernel Library Intel® optane™Intel® Xeon® Scalable processorsCloud
- 21. 0 1 2S Intel® Xeon®processor E5-2699 v4 2S Intel® Xeon® Platinum 8180 processor 21 QihooSearchEngine SearchEngine Application: Qihoo Search Engine (so.com) is the second largest Chinese search engine, and is widely used in PRC. Decompress is one of the most time-consuming modules in search engine. Potential Customer Benefits: Accelerate Qihoo Search Engine (so.com) decompress operation to reach breakthrough performance, and greatly improve the user experience. Performance Factors: Increased core count, higher frequencies and greater memory bandwidth of the Intel® Xeon® Platinum 8180 processor were key to the performance gain. “Decompress module is one of the critical modules for search engine overall performance. We are glad to see 56% performance improvement for our search engine decompress module on the latest on Intel® Xeon® Platinum 8180. This will greatly improve our production’s user experience.” Yang Kang, General Manager Of Search Business Unit “解压缩模块是决定搜索引擎性能的关键模块之一。我们很高兴的看到在新的Intel® Xeon® Platinum 8180 服务器上，我们搜索引擎的解压缩性能比上一代平台提升了56%。这将大大提升我们产品的用户体验。”, 杨 康, 搜索事业部总经理 Up to 56% faster Qihoo Search Engine (so.com)* increased performance with the Intel® Xeon® Platinum 8180 processor1 1 - Testing conducted on Qihoo Search Engine* software comparing Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4 Testing by Intel. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedPerformance http://so.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Workload: Qihoo Search Engine MB/second Intel® Xeon® Scalable processorsCloud
- 22. 0 1 2S Intel® Xeon®processor E5-2699 v4 2S Intel® Xeon® Platinum 8180 processor 22 TongTechTongLink/Q Application: TongLink/Q is a Queue middleware widely used in Tongtech’s banking system and highway management systems. Potential Customer Benefits: Support more transactions at the same time and save every transaction’s time to improve user’s experience. Performance Factors: Increased core count, higher frequencies and greater memory bandwidth of the Intel® Xeon® Platinum 8180 processor were key to the performance gain. “Taking technical advantage of the Intel® Xeon® Platinum 8180, we see a great performance boost on TongTech TongLink/Q compared with the previous generation processor. We will surely recommend this platform to our TongLink/Q customers.” Zou Qiang, Development manager “基于Intel® Xeon® Platinum 8180处理器的强大技术优势，东方通TongLink/Q与上一代平台相比有着显著的 性能提升，我们一定会向我们TongLink/Q的客户推荐这个平台。”，邹强，研发经理 TongTech TongLink/Q* increased performance with the Intel® Xeon® Platinum 8180 processor1 1 - Testing conducted on TongLink/Q* software comparing Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4 Testing by Intel. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedPerformance www.tongtech.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Workload: TongTech TongLink/Q middleware transactions per second (TPS) Up to 56% faster Intel® Xeon® Scalable processorsCloud
- 23. 0 1 Gen-to-Gen Intel® AVX-512 2SIntel® Xeon® processor E5 2699 V4 2S Intel® Xeon® Platinum 8180 Processor 2S Intel® Xeon® Platinum 8180 Processor w/Intel® AVX-512 23 LeCloudVideoTranscoding* Application: LeCloud Video Transcoding is to conduct video operations for its customer. We applied AVX-512 optimize on Intel® Xeon® Platinum 8180 Platform for the original function and achieved great performance speedup. Potential Customer Benefits: Reduced latency in video transcoding More fantastic and smoothing user experience for LeCloud customers Performance Factors: More cores and high memory bandwidth of the platform were key to the overall performance gain of 1.51X over previous processor generation1. Software optimizations leveraging Intel® AVX-512 delivered a 22% performance improvement1 over using Intel® AVX2 instructions, increasing the speed of the video transcoding process. “The Intel® Xeon® Platinum 8180 processor is a powerful platform for our transcoding, and certainly will improve the user experience.” Fu Qingming, LeCloud CTO “Intel® Xeon® Platinum 8180 为我们的视频转码提供了强大的底层平 台，必将大大提升我们的用户体验”，符庆明, 乐视云CTO ATAGLANCE Hardware: Intel® Xeon® 8180 Platinum Processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® Compiler Up to 51% faster LeCloud Video Transcoding* increased performance with the Intel® Xeon® Platinum 8180 processor 1 - Testing conducted on LeCloud* software comparing 2S Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4 Testing by Intel. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedPerformance www.lecloud.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Media Performance Metric: frame/second (gen-to-gen), second (AVX-512) Up to 22% faster w/ Intel AVX-512 Intel® AVX-512 More info Case Study Intel® Xeon® Scalable processorsCloud
- 24. 24 Citrixxenserver7.2* Citrix XenServer* isa leading virtualization management platform optimized for application, desktop and server virtualization infrastructures. “Citrix XenServer are pleased to add support for the Intel® Xeon® processor Scalable family in XenServer 7.2*, released in May 2017. This support enables both customers using XenServer* directly, and those using XenServer as a component of their XenApp*/XenDesktop* deployments to take advantage of these latest CPUs and the performance improvements they offer.” – Bob Ball, Principal Software Engineer, Citrix Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Processor Scalable Family www.citrix.com Up to 74% faster More info https://www.citrix.com/products/xenserver/ VirtualizationCloud
- 25.
- 26. Communications services 26 Intel®Xeon®ProcessorScalableFamily Performance Proof Points– Communications Services GENERATION- TO-GENERATION INTEL® QUICKASSIST TECHNOLOGY INTEL® OPTANE™ TECHNOLOGYPotential Customer Benefits: • Increased throughput for optimized service quality • Enhanced responsiveness for improved user experience • Accelerated processing to support novel services 1.9X 1 1.29X 1.56XUp to Up to Up to 1 – As demonstrated by proof points in this presentation Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others 1 1 Average Gen-to-Gen processor application speedup: 1.52X See slide 28 See slide 27 See slide 27
- 27. 1.41 2.21 1.29 0 1 2 Gen-to-Gen Intel® QuickAssist Technology 4SIntel® Xeon® processor E7-8890 v4 Intel® Xeon® Platinum 8180 Processor Intel® Xeon® Platinum 8180 Processor w/Intel® Optane™ Intel® Xeon® Platinum 8180 Processor w/Intel® QAT 27 AsiaInfoTelcoBSS comms Application: AsiaInfo is the largest BSS (Business Support System) provider in PRC, and its Telco BSS is a fundamental application for telecom carriers. Performance is critical for this compute-bound application. Potential Customer Impact: Accelerate customer-facing operations at peak loads to deliver a better user experience. Modernize your datacenter for powerful performance and maximum scalability. Efficient encryption/decryption of sensitive data with integrated Intel® QAT. Improve cost efficiency of operations with Intel® Optane™ SSDs. Optimizations / Platform Features: More cores and high memory bandwidth were key to the performance gain. Integrated Intel® QAT speeds up encryption of sensitive data by up to 29% (see chart).1 High-speed, low-latency access to data with Intel® Optane™ SSDs dramatically improves I/O efficiency (additional 56% faster; see chart).1 “The new Intel® Xeon® Scalable Processors demonstrates its technology advantage with a 41% performance improvement1 over the previous platform, as well as added benefits from Intel® QAT. This will provide a better user experience for our customers, while providing high reliability and high performance.” Jiang Mingjun, Product Manager, AsiaInfo “基于英特尔® 至强® 8180的服务器显示了其先进的技术：跟以往至强平台相比1.41倍 的性能提升， Intel® QAT也提供了额外的性能提升。它将给我们的客户带来一个高可 靠高性能的平台，为用户提供更好的用户体验。”，姜明俊, 亚信产品经理 ATAGLANCE Hardware: 4S Intel® Xeon® Platinum 8180 Processor Intel® Optane™ Solid State Drive DC P4800X Platform Features: Integrated Intel® Quick Assist Technology (intel® QAT) Intel® Virtual Raid on CPU AsiaInfo Telco BSS* increased performance with the 4S Intel® Xeon® Platinum 8180 processor, Intel® Optane™ and SSD and integrated Intel® QAT 1 - Testing conducted on AsiaInfo Telco BSS* software comparing 4S Intel® Xeon® Platinum 8180 processor to 4S Intel® Xeon® Processor E7-8890 v4. Testing by AsiaInfo and Intel. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedPerformance http://www.asiainfo.com.cn Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors More info See the Solution Brief and Infographic INTEL® VROC Performance Metric: K-TPS (k-transaction/second) Up to 1.56X w/Intel® Optane Intel® optane™Intel® QAT4 socket
- 28. 0 1 2S Intel® Xeon®processor E5-2699v4 2S Intel® Xeon® Platinum 8180 processor 0 1 2 2S Intel® Xeon® processor E5-2699v4 2S Intel® Xeon® Platinum 8180 processor 28 eBriskVideoHEVCCodec* Application: eBLive A-5000 is an HEVC encoder that ships in two modes: eBLive A-5000-HQ: High-Quality (HQ) real-time encoder @ bit rates of <= 15 Mbps for high-quality video broadcast applications eBLive A-5000-HD: High Density (HD) real-time encoder @ bit rates of <=15 Mbps for real-time video streaming applications eBLive A-5000 can use all available cores of the listed processors, using at most 15GB of memory footprint per channel. Potential Customer Benefits: The Intel® Xeon® 8180 Platinum processor allows the customer to support either two 4kp60/10bit/HDR channels using eBrisk’s eBLive A-5000-HQ HEVC encoder or four 4kp60/10bit/HDR channels using eBrisk’s eBLive A-5000-HD HEVC encoder on a compact 1RU solution. Performance Factors: By taking advantage of the Intel® Xeon® 8180 Platinum processor architecture, higher memory bandwidth, and advanced caching algorithms, the eBLive A-5000-HQ experienced a computational performance gain of 1.9:1 relative to the Broadwell Intel® Xeon® processor E5-2699 v4. ATAGLANCE Hardware: Intel® Xeon® 8180 Platinum processor Platform Features: More cores Higher memory bandwidth Software Tools/Libraries: Intel® Compiler Intel® VTune™ Amplifier 1 - Testing conducted on eBrisk video software comparing 2S Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4. Testing by eBrisk. For complete testing configuration details, see the Configurations Details section, slide 135. www.ebriskvideo.com Performance Metric: Channels per Socket eBLive A 5000 increased channel density with the Intel® Xeon® Platinum 8180 processor NormalizedChannelsNormalizedPerformance eBLive A 5000* increased performance with the Intel® Xeon® Platinum 8180 processor eBLive A-5000-HQ eBLive A-5000-HD Performance Metric: Channels per Socket Up to 90% faster Up to 70% faster Up to 2x more channels Up to 2x more channels eBLive A-5000-HQ eBLive A-5000-HD Intel® Xeon® scalable Processors Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Video processingcomms
- 29. 29 SurfCommunicationsSolutionsSurf-HMP* Application: SURF-HMP is acomprehensive multimedia processing platform with next-gen capabilities enabling ultra- HD video, voice and signal processing. SURF-HMP can be embedded in your solution on premise, in the Cloud, or in a NFV infrastructure. Potential Customer Benefits: The Intel® Xeon® Platinum 8168 processor produced an up to 1.37X1 performance boost per core compared to the Intel® Xeon® processor E5-2690 v3 which means customers can gain 37% more H.264 full HD video channels per core. The Intel® Xeon® Platinum 8168 processor produced an overall system level improvement of up to 2.75X1 for full H.264 HD video channel density compared to the Intel® Xeon® processor E5-2690 v3, which reduces TCO (less servers required to produce the desired production), and eliminates the needs for a discrete GPU. Performance Factors: The Intel® Xeon® Platinum 8168 processor provides twiece as many cores (24) compared to the Intel® Xeon® processor E5-2690 v3 (12), more memory channels (6 vs 4), and faster memory support (2666MHZ vs 2133MHZ). Faster communications between cores (Intel® Mesh Architecture). “New Intel’s Purley system provides Surf the ability to deliver far better performance, outperform current offering with discrete GPU“ Niv Kagan, VP R&D, Surf Communications Solutions 1 - Testing conducted on ISV* software comparing 2S Intel® Xeon® Platinum 8168 processor to 2S Intel® Xeon® Processor E5-2690 v3, H.264 full HD video channels per core workload. Testing done by Surf Communications Solutions. For complete testing configuration details, see the Configurations Details section, slide 135. www.surfsolutions.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable processors 0 1 2S Intel® Xeon® processor E5-2690 v3 2S Intel® Xeon® Platinum 8168 processor 0 1 2 2S Intel® Xeon® processor E5-2690 v3 2S Intel® Xeon® Platinum 8168 processor Channel Density per System improved by up to 2.75X SURF-HMP increased channel density with the Intel® Xeon® Platinum 8168 processor NormalizedPerformanceNormalizedPerformance SURF-HMP increased performance with the Intel® Xeon® Platinum 8180 processor Channel Density per Core improved by up to 1.37X Video processingcomms multi-gen
- 30. 0 1 Main Profile Main10 Profile 2S Intel® Xeon® processor E5-2699 v4 2S Intel® Xeon® Platinum 8180 Processor 30 Beamr5HEVCEncoder* Application: Beamr 5 is a high-performance, broadcast quality software encoder that delivers real time HDR and SDR HEVC encodings Potential Customer Benefits: 2x HEVC real time encoding density: encode HDR 10-bit 2160p60 on one (1) socket of Platinum 8180 versus dual socket E5-2699 v4 Improve HEVC visual quality. Deploy more intensive presets (from “superfast” to “medium”) to achieve better quality without loss of throughput. Deliver complex workflows in less time, with less equipment. Performance Factors: Intel® Xeon® Platinum 8180 processor provides more cores (28) compared to the previous processor generation (22), more memory channels (6 vs 4), and faster memory support (2666MHZ vs 2400MHZ). Faster communications between cores (Intel® Mesh Architecture). Beamr 5* increased performance with the Intel® Xeon® Platinum 81801 1 - Testing conducted on Beamr* software comparing 2S Intel® Xeon® Platinum 8180 processor to Intel® Xeon® processor E5-2699 v4. Testing by Beamr. For complete testing configuration details, see the Configurations Details section, slide 135. www.beamr.com Workload: 4:2:0 8bpp and 10bpp, and 4:2:2 10 bpp clips were encoded on the respective processors, and no other applications running. Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others ATAGLANCE Hardware: Intel® Xeon® 8180 Platinum processor Platform Features: More cores More memory channels Software Tools/Libraries: Intel® Inspector Intel® VTune™ Amplifier NormalizedPerformance Up to 75% faster Up to 63% faster Intel® Xeon® scalable ProcessorsVideo processingcomms
- 31. 0 1 2S Intel® Xeon®processor 2699 v3 Intel® Xeon® Gold 6150 processor 31 SandvineVirtualSeries Application: Supporting industry and open standards, the Sandvine Virtual Series delivers Sandvine’s powerful Policy Engine, PCEF, and PCRF as an elastically-scalable virtual network function for SDN- and NFV-ready networks. Potential Customer Benefits: Communications service providers (CSPs) will now need to utilize fewer vCPUs in their virtual workload environment to achieve a given level of throughput and to enable Sandvine’s network policy control platform at line rate speeds. Performance Factors: Intel® Xeon® Scalable processor microarchitecture – including more cores and memory bandwidth compared to previous processor generations – provided an up to 60% performance improvement1 for Sandvine’s Virtual Series. “Without any optimization to our PTS Virtual Series software, the Intel® Xeon® Scalable processor architecture will provide our customers with the ability to do significantly more with the Sandvine platform without the need foradditional vCPUs. Later this year when we update our Virtual Series products to take advantage of the Intel® AVX-512 instruction set, we will be able to enhance our cyber security solutions with innovative new features based on deep machine learning and neural-network acceleration.” - Don Bowman, Sandvine CTO Up to 60% better throughput Sandvine Virtual Series increased performance with the Intel® Xeon® Gold 6150 Processor 1 - Testing conducted on Sandvine PTS Virtual Series comparing Intel® Xeon® E5-2699v3 processor with Intel® Xeon® Gold 6150 processor. Testing done by Sandvine. For complete testing configuration details, see the Configurations Details section, slide 135. Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable processors SKUs & Adjacencies Relevant platform technologies www.sandvine.com ATAGLANCE Hardware: Intel® Xeon® Gold 6150 processor Intel® 82599ES 10 Gigabit Ethernet Controller Software Tools/Libraries: Data Plane Development Kit NormalizedPerformance Workload: Packet inspection NFVcomms
- 32. 0 1 Live SD H.264Live HD HEVC 2S Intel® Xeon® processor E5-2699 v4 2S Intel® Xeon® Platinum 8168 Processor Workload: KEEPIXO number of live TV services 32 KEEPIXOGENOVALIVE* Application: Genova Live is a real-time H.264 and HEVC video transcoder for Internet TV. This is a core bound application. Potential Customer Benefits: Higher density and lower hardware cost per TV service Lower bitrate (by 6%), translating to lower network bandwidth usage and lower cost, for the same video quality and density Performance Factors: Transcode up to 55% more live HD HEVC TV services or up to 43% more live SD H264 using the Intel® Xeon® 8168 Platinum processor compared to previous generation, due to increased number of cores. By recompiling the source code with Intel compiler and enabling AVX-512 instructions a 7% performance gain was measured for SD H264 transcoding. ATAGLANCE Hardware: Intel® Xeon® 8168 Platinum processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries Intel® Compiler ICC 2017 Intel® Integrated Performance Primitives (Intel® IPP) Up to 43% more live TV services Genova Live* increased performance with the Intel® Xeon® Platinum 8168 processor1 1 - Testing conducted on Keepixo* software comparing Intel® Xeon® Platinum 8168 processor to 2S Intel® Xeon® Processor E5-2699 v4 T. Testing by Keepixo. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedPerformance www.keepixo.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable ProcessorsVIDEO 1451 73 9 Intel® AVX-512 Up to 55% more HD services comms
- 33. 0 1 2S Intel® Xeon®processor 2699 v4 2S Intel® Xeon® Platinum 8168 processor 47 fps 71 fps NormalizedPerformance 33 EricssonMediaFirstVideoProcessingEncodingLive* VIDEO Processing Application: Ericsson MediaFirst Video Processing Encoding Live is a software transcoder which encodes in real time UHD video: 4K60 HEVC 10 bits with HDR. This workload takes benefit of the multicore architecture of the vectorised integer instructions and is dependent on the available memory bandwidth. Potential Customer Impact: Ability to deliver broadcast quality in real-time (>60 fps). Meet performance requirements with a 2S Intel® Xeon® Scalable Processors system, where a 4 socket server was needed with the previous generation Intel® Xeon® Processor E5 v4 Family. Performance Factors: Increased core count, higher frequencies and greater memory bandwidth of the Intel® Xeon® Platinum 8168 processor were key to the performance gain. Application also benefits from improved cache management of the processor. ATAGLANCE Hardware: Intel® Xeon® Platinum 8168 processor Software Tools/Libraries : Intel® Compiler ICC 2017 Intel® Integrated Performance Primitives (Intel® IPP) Intel® VTune™ Amplifier Ericsson MediaFirst* increased performance with the Intel® Xeon® Platinum 8168 processor1 1 - Testing conducted on Ericsson* software comparing Intel® Xeon® Platinum 8168 processor to 2S Intel® Xeon® Processor E5-2699 v4. Testing by Ericsson. For complete testing configuration details, see the Configurations Details section, slide 135. www.ericsson.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Workload: Ericsson MediaFirst Video Processing UHD HEVC transcoding frames per second (FPS) comms Up to 51% faster
- 34. 0 1 2S Intel® Xeon®processor E5-2680 v4 2S Intel® Xeon® Platinum 8170 processor Workload: Gigabits per sec of IPsec traffic per core 34 6windSite-to-SiteIPsecVPN* Application: Site-to-site IPsec VPN provides high-speed, private and confidential communications without the need for expensive WAN connections. In addition, customers leverage low cost servers that can provide the same high-speed communications instead of expensive, proprietary hardware solutions. Potential Customer Benefits: Scalable (to > 40 Gbps) IPsec solution. Software Portability (processor independence). Performance Factors: Up to 1.5X performance improvement1 with the Intel® Xeon® Platinum 8170 processor and Intel® AES-NI (with Galois/Counter Mode, aka GCM), measuring encrypted traffic per second, compared to the previous processor generation. ATAGLANCE Hardware: Intel® Xeon® Platinum 8170 processor Platform Features: Intel® Advanced Encryption Software – New Instructions (Intel® AES-NI) Software Tools/Libraries: Data Plane Development Kit Up to 50% faster 6WIND Turbo IPsec increased performance with the Intel® Xeon® Platinum 8170 processor 1 - Testing conducted on 6WIND software comparing Intel® Xeon® Platinum 8170 processor @ 2.1 GHz to Intel® Xeon® Processor E5-2680 v4 @ 2.4 GHz. Testing by 6WIND. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedPerformance www.6WIND.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others “At 6WIND, we are pleased with the increase in performance of Intel's new platform that translates directly to an increase in performance of our security software. This pairing allows standard servers to rival specialized hardware systems so that customers never have to sacrifice performance for cost. Together, 6WIND and Intel provide a cost-effective alternative to expensive, specialized hardware for high performance site-to-site VPN solutions used to secure data center to data center data communications.” Eric Carmès, CEO and Founder of 6WIND Security Intel® Xeon® scalable Processors Intel® AES-NI comms
- 35. 0 1 2S Intel® Xeon®processor 2699 v4 2S Intel® Xeon® Platinum 8180 processor Workload: Frames per second for encoding 8-bit 1080p 35 x264Open-sourceAVCEncoder Application: Open source AVC encoder widely used in cloud and comms applications. Potential Customer Benefits: 44% higher throughput1 for offline file-based encoding. 2X density per socket improvement1 for 1080p60 live encoding. Performance Factors: Increased thread count per socket translates to higher density per-socket. Generations of SIMD optimizations (including Intel® AVX- 512) contributed to up to 2X performance gains. “Significantly higher performance and more cores per socket deliver up to 2x AVC encoding density!” Tom Vaughan, VP & GM - Video at MulticoreWare ATAGLANCE Hardware: Intel® Xeon® 8180 Platinum processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) 44% faster x264 performance increase with Intel® Xeon® Platinum 8180 processor 1 - Testing conducted on x264 comparing 2S Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® processor E5-2699 v4 Testing by MulticoreWare Inc and Intel. For complete testing configuration details, see the Configurations Details section, slide 135. More info http://www.videolan.org/developers/ x264.html NormalizedPerformance www.multicorewareinc.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others 0 1 2 2S Intel® Xeon® processor 2699 v4 2S Intel® Xeon® Platinum 8180 processor Workload: Channels per socket for 8-bit 1080p60 2X channels per socket x264 per-socket 1080p60 density doubles with Intel® Xeon® Platinum 8180 processor Intel® AVX-512 NormalizedDensity Intel® Xeon® scalable Processor FamilyVideo processingcomms
- 36. 0 1 2S Intel® Xeon®processor 2699 v4 2S Intel® Xeon® Platinum 8180 processor Workload: Frames per second for 10-bit 1080p & 4K 36 x265Open-sourceHEVCEncoder Application: Open source HEVC encoder, widely used in comms and cloud applications for video delivery. Application is sensitive to both core and memory performance and employs extreme levels of thread parallelism and optimizations. Potential Customer Benefits: 56% quicker1 turn-around for offline file-based 1080p and 4K encoding. 1.33X improved1 density per socket for 1080p60 live encoding. Enables new extremely low-latency applications. Performance Factors: Increased thread count per socket translates to higher density per-socket. “56% higher x265 performance, combined with higher core counts delivers 2x performance per server!” Tom Vaughan, VP & GM - Video at MulticoreWare 56% faster x265 encoding performance improvement with Intel® Xeon® Platinum 8180 processor 1 - Testing conducted on x265 comparing 2S Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4 Testing by MulticoreWare and Intel. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedPerformance www.multicorewareinc.com 0 1 2S Intel® Xeon® processor 2699 v4 2S Intel® Xeon® Platinum 8180 processor Workload: Channels per socket for 10-bit 1080p60 x265 increased density for live 1080p60 10-bit broadcast with Intel® Xeon® Platinum 8180 processor NormalizedDensity 1.33X channels per socket Intel® Xeon® scalable Processors Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Video processingcomms
- 37. 0 1 2S Intel® Xeon®processor 2699 v4 2S Intel® Xeon® Platinum 8180 processor Workload: BD-rate when sustaining live 10-bit 4K60 37 UHDkit:ACommercialAVC/HEVCEncoder Application: Enables commercial AVC/HEVC SW encoding use-cases. Runs multiple x264/x265 instances for live AVC/HEVC encoding. Improves quality and performance for file-based and live encoding. Potential Customer Benefits: Significantly higher real-time HEVC video encoding quality or achieve 20% lower bit rates at identical quality. Live 4K60 encoding 10-bit 4:2:2 now practical in software! Performance Factors: Benefits from improved single-threaded and memory performance. “The new Intel Xeons deliver significantly higher quality and compression efficiency for live 4K 60P 10 bit HEVC video encoding!” Tom Vaughan, VP & GM - Video at MulticoreWare UHDkit live 4K60 10 bit HEVC video encoding quality improved with Intel® Xeon® Platinum 8180 processor 1 - Testing conducted on UHDkit comparing 2S Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4 Testing by MulticoreWare and Intel. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedResults-Bandwidthconsumed www.multicorewareinc.com Delivers 20% higher live 4K60P HEVC compression efficiency Intel® Xeon® scalable Processors Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Video processingcomms
- 38. 0 1 4S Intel® Xeon®processor E7-8890 v4 4S Intel® Xeon® Platinum 8180 processor 38 ShenzhenTYDICOCSonlinechargingsystem* Application: Shenzhen TYDIC is a leading company in the fields of providing business support systems for domestic telecommunications industry. TYDIC OCS is a telecom distributed online charging system. Potential Customer Benefits: Accelerate system throughput to deliver better customer services and reduce the response time. Performance Factors: Increased core count, higher frequencies and greater memory bandwidth of the Intel® Xeon® Platinum 8180 processor were key to the performance gain. “OCS is the most important part of TYDIC Operation Support System. The Intel® Xeon® Platinum 8180 processor will help to provide better performance and reduce the operation latency.” Liu Yu, director of Billing product center “在线计费系统是电信运营支撑业务中最核心的模组。Intel® Xeon® 8180 Platform的强大性能将有助 于为我们的客户提供更好的系统性能，减少业务的处理时间”，刘羽，Billing产品中心总监 TYDIC online charging system increased the performance by 1.41x with the Intel® Xeon® Platinum 8180 processor 1 - Testing conducted on TYDIC* software comparing 4S Intel® Xeon® Platinum 8180 processor to 4S Intel® Xeon® Processor E7-8890 v4 Testing by Intel and TYDIC. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedPerformance www.tydic.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Up to 41% faster 4 socket Workload: TYDIC online charging system comms
- 39. 0 1 2S Intel® Xeon®processor 2699 v4 Intel® Xeon® Platinum 8168 processor 39 PexipInfinity2.6* Application: Pexip Infinity is video communications software that organizations can deploy across their IT infrastructure, in a public or private cloud or any hybrid combination – so that every individual can have their own personal high-quality video, audio and mobile meeting experience. Potential Customer Benefits: Higher number of 1080HD video streams that Intel® Xeon® 8168 Platinum can handle (2x61 compared to 2x45 on E5-2699 v4, VP9) enables more streams per server which keeps TCO down. Performance Factors: Manual and compiler Intel® AVX-512 enabling. “The VP9 video codec with superblocks of 64 x 64 pixels is particularly well suited to the Intel AVX-512 instruction set since a 512 bit registers can hold exactly 64 pixels. At similar visual quality, our VP9 implementation is roughly 3 times faster than the LibVPX implementation that is not yet adopted to AVX512.” - Lars Petter Endresen, PhD, Principal Performance Engineer, Pexip “The Intel Xeon Scalable processors with Intel AVX-512 provides up to almost 50% performance increase for VP9 encoding and decoding [compared to the previous generation Broadwell CPUs]. This translates to a dramatic reduction in TCO for our customers whether they deploy in an IaaS cloud or on-premises, and will enable very high quality enterprise video communication at scale.” – Håkon Dahle, CTO, Pexip ATAGLANCE Hardware: Intel® Xeon® Platinum 8168 processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX- 512) Software Tools/Libraries: Intel® Parallel Studio XE 2018 Up to 35% faster Pexip Infinity* increased performance with the Intel® Xeon® Platinum 8168 processor1 1 - Testing conducted on Pexip* software comparing Intel® Xeon® Platinum 8168 processor to 2S Intel® Xeon® Processor E5-2699 v4. Testing by Pexip. For complete testing configuration details, see the Configurations Details section, slide 135. MORE info https://www.webmproject.or g/code/vp9/ NormalizedPerformance www.pexip.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® scalable Processor Family Intel® AVX-512 Workload: Pexip’s Infinity 2.6 video conferencing handling maximum number of 1080HD video streams (withVP9 codec ) Video processingcomms
- 40. This ENTIRE deckis embargoed until July 11, 2017, 9:30am PDST Enterpriseapplications
- 41. 41 Intel®Xeon®ProcessorScalableFamily Performance Proof Points– Enterprise INTEL® OPTANE™ TECHNOLOGY MULTI- GENERATION Potential Customer Benefits: • Continuous real-time analytics for actionable insight • Higher throughput for improved user experiences • Accelerated response times to foster innovation Enterprise Up to 1.35X Up to 4X2.14X GENERATION- TO-GENERATION Up to 1 1 – As demonstrated by proof points in this presentation Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others 1 1 Average Gen-to-Gen processor application speedup: 1.54X Average Multi Gen processor application speedup: 2.27X See slide 42 See slide 61 See slide 41
- 42. 0 2 4 2S Intel® Xeon®E5-2697 v3 + Aerospike 3.6.4 2S Intel® Xeon® E5-2699 v4 + Aerospike 3.12.1 2S Intel® Xeon® Platinum 8180 + Aerospike 3.12.1 Up to 2.8X faster 42 AEROSPIKE* Enterprise Application: Aerospike is a Hybrid Memory Architecture database optimized for flash and SSD. It is the database of choice for systems of engagement (SoE). Potential Customer Benefits: Process more transactions at sub-millisecond response times. Deliver a better user experience. Performance Factors: Software enhancements in the latest version of Aerospike Server 3.12.1, such as “auto- tuning” to automatically bind processes and network interrupts to CPU cores. Uniquely designed to take advantage of more cores & higher memory bandwidth of the latest Intel® Xeon® Platinum 8180 processor. Up to 4X faster Aerospike Server* increased performance with Aerospike 3.12.1 and the Intel® Xeon® Platinum 8180 processor1 1 - Testing conducted on Aerospike* software comparing Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2697 v3 and to 2S Intel® Xeon® Processor E5-2699 v4 Testing done by Intel. For complete testing configuration details, see the Configurations Details section, slide 135 more INFO Solution Brief 2x2 NormalizedPerformance www.aerospike.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors multi-genDatabase Workload: Simulated 95%/5% read/update ratio, TPS 1.42X
- 43. 0 1 2 2S Intel® Xeon®E5-2699 v4 2S Intel® Xeon® Platinum 8168 processor Workload summary: TPC-H like; queries per second Up to 1.51X faster 43 EXASOL* Application: The world’s fastest (according to TPC-H benchmarks), highly parallel, scalable in-memory clustered database. Ideally suited for real-time big data reporting, analysis and advanced analytics. Workload – custom TPC-H like. Potential Customer Benefits: Extra performance gain, more concurrent users. Performance Factors: In-depth and long-term multi-core optimization. Efficient utilization of new enhanced processor cores and cache. “The new Intel Xeon Platinum processor delivers yet even more analytic performance. We see up to 1.51x for 2 socket and up to 2.14X for 4 socket extra performance vs. previous generations of corresponding Intel® Xeon® processors for our latest EXASOL database in multi-user tests. Running on the new Intel® Xeon® Platinum processor-based servers allows our users to benefit from a whole new level of performance in our innovative in-memory DB technology.” André Hacker, Head of R&D, EXASOL AG EXASOL* increased performance with the 2S Intel® Xeon® Platinum 8168 processors1 1 - Testing conducted on Exasol* software comparing 4s Intel® Xeon® Platinum 8180 processor to 4S Intel® Xeon® Processor E7-8890 v4 and 2s Intel® Xeon® Platinum 8168 processor to 2S Intel® Xeon® Processor E5-2699 v4 Testing done by ISV/Intel. For complete testing configuration details, see the Configurations Details section, slide 135 NormalizedPerformance www.exasol.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors 0 1 2 4S Intel® Xeon® E7-8890 v4 4S Intel® Xeon® Platinum 8180 processor Workload summary: TPC-H like; queries per second Up to 2.14X faster EXASOL* increased performance with the 4S Intel® Xeon® Platinum 8180 processors1 In-memory database 4 socket Enterprise
- 44. 0 1 2 2S Intel® Xeon®processor E5-2699 v4 Intel® Xeon Platinum 8180 w/ Intel® Optane™ Workload: SAS 9.4 Mixed Analytics; Metric: SAS Real Time 44 SAS9.4Mixedanalytics Application: SAS applications provide an integrated environment for predictive and descriptive modeling, data mining, text analytics, forecasting, optimization, simulation, experimental design and more. The SAS Mixed Analytics workload simultaneously generates extremely high and prolonged CPU utilization, and heavy large-block sequential IO demand in addition to small random reads and writes. The workload simulates a typical SAS analytics environment with 30 users running parallel analytics tasks. Potential Customer Benefits: When coupled with the Intel® Optane™ SSD DC P4800X, the Intel® Xeon® Platinum 8180 processor improves throughput, allowing customers to run more complex analyses, in aggregation, in less time and across larger data sets, to gain deeper insight into critical business. Performance Factors: Greater number (27% more) cores and threads, and 50% more memory bandwidth than the previous generation Intel® processor. Improved cache hierarchy. The Intel® Optane™ SSD DC P4800X delivered a 27%1 performance gain compared to the Intel® P3700 SSD. ATAGLANCE Hardware: Intel® Xeon® Platinum 8180 processor Intel® Optane™ SSD DC P4800X Software Tools/Libraries : Intel® Parallel Studio XE Intel® Math Kernel Library Up to 2X faster SAS 9.4* increased performance with the Intel® Xeon® Platinum 8180 processor and the Intel® Optane™ SSD DC P4800X1 1 - Testing conducted on SAS 9.4* software comparing Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4 Testing done by Intel. For complete testing configuration details, see the Configurations Details section, slide 135 more info See the Solution Brief and Infographic NormalizedPerformance www.sas.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Intel® optane™ Enterprise analytics
- 45. 0 1 2 4S Intel® Xeon®processor E7-4890 v2 4S Intel® Xeon® processor E7-8890 v4 4S Intel® Xeon® Platinum 8180 processor 1.71X faster Workload BDInsights (3TB, 12 users, intermediate & complex queries) 45 IBM®DB2 ENTERPRISE Application1: DB2 is IBM’s flagship database product which supports in- memory column store tables for analytics workloads. The proprietary IBM Big Data Insights Workload (BDInsights) is a multi-user data warehousing workload based on a retail environment. The workload as configured uses a 3TB scale factor, 12 concurrent users, and a mix of intermediate and complex analytics queries, and is compute-bound in the tested configuration. Potential Customer Impact: Faster response time for analytic queries Higher throughput to support more concurrent users for customers. Performance Factors: Greater number of cores/threads and improved memory bandwidth of the Intel® Xeon processor Scalable Family Intel® C++ Compiler ATAGLANCE Hardware: Intel® Xeon® Platinum 8180 processor Software Tools/Libraries: Intel® C++ Compiler 2.53X faster IBM® DB2 11.1.1 increased performance with the Intel® Xeon® Platinum 8180 processor1 1 - Testing conducted on IBM DB2* software comparing Intel® Xeon® Platinum 8180 processor to 4S Intel® Xeon® Processor E7-4890 v2 and E7-8890 v4. Testing performed by Intel® and IBM®, April/May 2017. BASELINE: 4S Intel® Xeon® processor E7-4890 v2, 2.8GHz, 15 cores, turbo on, HT on, BIOS 38.R02, 1.5TB total memory, 96 slots / 16GB / 1066 MT/s / DDR3 DIMM, IBM XIV Storage (132 10K RPM disks), SuSE Enterprise Linux* 11.3 kernel 3.0.101-0.47.71. NEXT GEN: 4S Intel® Xeon® processor E7-8890 v4, 2.2GHz, 24 cores, turbo on, HT on, BIOS 335.R00, 1.5TB total memory, 96 slots / 16GB / 1600 MT/s / DDR4 LRDIMM, 1 x 800GB, Intel® SSD DC S3700, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.16.1.el7.x86_64. NEW: Intel® Xeon® Platinum 8180 processor, 2.5GHz, 28 cores, turbo on, HT on, BIOS 119.R05, 1.5TB total memory, 48 slots / 32GB / 2677 MT/s / DDR4 LRDIMM, 1 x 800GB, Intel® SSD DC S3700, Red Hat Enterprise Linux* X.X kernel 3.10.0- 514.16.1.el7.x86_64. More info • See the Solution Brief and Infographic • 2x2 NormalizedPerformance www.ibm.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors multi-gen DATABASE 4 socket
- 46. 0 1 2S Intel® Xeon®processor 2699 v4 2S Intel® Xeon® Platinum 8180 processor BDInsights (300GB analytics workload, 12 users) 46 IBM®DB2 Application1: DB2 is IBM’s flagship database product which supports in- memory column store tables for analytics workloads. The proprietary IBM Big Data Insights Workload (BDInsights) is a multi-user data warehousing workload based on a retail environment. The workload as configured uses a 300GB scale factor and 12 concurrent users. It is compute-bound in the tested configuration. Potential Benefits: Faster response time for analytic queries Higher throughput to support more concurrent users for customers. Performance Factors: Greater number of cores/threads and improved memory bandwidth of the Intel® Xeon processor Scalable Family Intel® C++ Compiler ATAGLANCE Hardware: Intel® Xeon® 8180 Platinum processor Software Tools/Libraries: Intel C++ Compiler 55% faster IBM DB2* 11.1.1 increased performance with the Intel® Xeon® Platinum 8180 processor1 MORE INFO https://www.ibm.com/analytic s/us/en/technology/db2/ NormalizedPerformance www.ibm.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors 1 - Testing conducted on IBM DB2* software comparing 2S Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4. Testing performed by Intel® and IBM®, April/May 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo on, HT on, BIOS 0271.R00, 256GB total memory, 16 slots / 16GB / 2133 MT/s / DDR4 LRDIMM, 1 x 800GB, Intel® SSD DC S3700, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.16.1.el7.x86_64. NEW: 2S Intel® Xeon® Platinum 8180 processor, 2.5GHz, 28 cores, turbo on, HT on, BIOS 00.0412, 192GB total memory, 12 slots / 16GB / 2677 MT/s / DDR4 LRDIMM, 1 x 800GB, Intel® SSD DC S3700, Red Hat Enterprise Linux* X.X kernel 3.10.0-514.16.1.el7.x86_64. DATABASEEnterprise
- 47. 0 1 SHA AES RSA 2SIntel® Xeon® processor E5-2699 v4 2S Intel® Xeon® Platinum 8180 processor 47 IBM®GlobalSecurityKIT Application: IBM® Global Security Kit (GSKit) provides libraries and utilities to enable SSL or TLS communication for IBM products, including DB2* and Websphere*. GSKit is a component and not a stand-alone product. GSKit support and updates are provided as part of other products' support and updates. The workload is compute bound and tests the speed of cryptographic operations including SHA, AES, and RSA. Potential Customer Benefits: Lowering the overhead for cryptographic operations like AES encryption lowers the barrier for adoption, delivering a better user experience and security. Performance Factors: IBM® GSKit leverages Intel® Advanced Encryption Standard New Instructions (Intel® AES-NI) to accelerate AES encryption. Up to 76% faster IBM GSKit* increased performance with the Intel® Xeon® Platinum 8180 processor1 1 - Testing conducted on IBM GSKit* software comparing Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4 Testing by Intel® and IBM®. For complete testing configuration details, see the Configurations Details section, slide 135 NormalizedPerformance www.ibm.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Up to 49% faster Up to 34% faster Workload: ICC 8.5, GSKit performance test securityEnterprise Intel® AES-NI
- 48. 0 1 2S Intel® Xeon®processor 2699 v4 2S Intel® Xeon® Platinum 8180 processor Workload: BDInsights 300GB analytics workload (12 users) 48 IBM®dashDBLocal Application: IBM® dashDB Local is a software-defined data warehouse for private clouds and virtual private clouds that support Docker container. The IBM® Big Data Insights Internal Heavy Multiuser Workload (BDInsights) is a multi-user data warehousing workload based on a retail environment. The workload is compute bound. Potential Customer Benefits: Faster response time for analytic queries and higher throughput to support more concurrent users for customers. Performance Factors: Intel® C++ Compiler Intel® AVX2 Up to 60% faster IBM® dashDB Local* increased performance with the Intel® Xeon® Platinum 8180 processor1 1 - Testing conducted on IBM® dashDB Local* software comparing Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4. Testing by Intel® and IBM®. For complete testing configuration details, see the Configurations Details section, slide 135 More info https://www.ibm.com/us- en/marketplace/ibm-dashdb- local NormalizedPerformance www.ibm.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable ProcessorsdatabaseEnterprise
- 49. 0 1 2S Intel® Xeon®processor E5-2699 v4 2S Intel® Xeon® Platinum 8180 processor Performance Metric: Records processed per second 49 IBM® AlgoFinancialModeler® ENTERPRISE Application: IBM Algo Financial Modeler is an application that models financial contracts. The complex stochastic actuarial model used for these tests has 1000 scenarios, where each record is valued over a 120 year period using 1 month projection steps. Potential Customer Benefits: Many client datasets have more than a million records, thus throughput is important. In this model, run times scale linearly with the number of records. The calculation is not I/O bound and makes full use of all allocated cores. Performance Factors: The Intel® VTune™ Amplifier performance profiler highlighted model and generator improvements that helped increase throughput on all machines. “New industry regulations require an increase in model complexity; the use of the Intel® Xeon® Platinum 8180 processor helps to meet their obligations.” ATAGLANCE Hardware: Intel® Xeon ® Platinum 8180 processor Software Tools/Libraries: Intel® VTune™ Amplifier 49% faster IBM Algo Financial Modeler* increased performance with the Intel® Xeon® Platinum 8180 processor1 1 - Testing conducted on IBM Algo Financial Modeler* software comparing Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® processor E5-2699 v4 Testing done by IBM. For complete testing configuration details, see the Configurations Details section, slide 135. More INFO www.ibm.com/software/prod ucts/en/algofinamode NormalizedPerformance www.ibm.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable ProcessorsRISK MANAGEMENT
- 50. 0 1 2S Intel® Xeon®processor 2699 v4 Intel® Xeon® Platinum 8180 processor Performance Metric: Throughput (queries per second) 50 IBM®INFORMIXWAREHOUSEACCELERATOR Application: IBM® Informix Warehouse Accelerator is an in-memory database accelerator for IBM Informix database server, which includes deep columnar and compression technology. Workload comprises of analytical queries on a set of unstructured sensor time series data. Potential Customer Benefits: Deliver higher throughput on analytical queries for customers. Performance Factors: Multi-core and vector optimized algorithms paired with massive parallelism, exploit high core count Intel® Xeon® processors with Single Instruction, Multiple Data (SIMD) technology. 47% faster IBM IWA* increased performance with the Intel® Xeon® Platinum 8180 processor1 1 - Testing conducted on IBM Informix Warehouse Accelerator* software comparing Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4. Testing done by IBM. For complete testing configuration details, see the Configurations Details section, slide 135 NormalizedPerformance www.ibm.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable ProcessorsanalyticsEnterprise
- 51. 0 1 vNAS 1.5.86 withoutencryption vNAS 1.5.86 with encryption Workload: vNAS 1.5.86 with Database benchmark. 51 COMPUVERDEvNAS* Application: Compuverde vNAS is a high performance scale out cloud storage solution optimised for flash storage and designed for linear scalability. Potential Customer Benefits: Full AES encryption for data at rest. With near zero overhead maintaining high throughput. Performance Factors: Leveraged open source Intel® ISA-L Crpyto library (using Bitlocker* XTS-AES 256). Intel® AES-NI and Intel® AVX-512 accelerated crypto functions1. Intel® SSD DC P3700 as write cache to further accelerate workload performance. “The performance in the encryption routines is very good; we don’t notice any performance degradation when it is activated” – Christian Melander, Senior Software Engineer, Compuverde ATAGLANCE Hardware: Intel® Xeon® Platinum 8180 Processor Intel® Solid State Drive Data Center P3700 Platform Features: Intel® Advanced Encryption Standard New Instructions (Intel® AES-NI) Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® Intelligent Storage Acceleration Library (Intel® ISA-L) Intel® AVX-512 Matchedperformance Using Intel® ISA-L with newest generation Xeon® Platinum 8180 Processor encrypting/decrypting data to/from storage disks 1 - Testing conducted on vNAS* software using Xeon® Platinum 8180 processor. Testing by Compuverde and Intel. For complete testing configuration details, see the Configurations Details section, slide 135 NormalizedPerformance www.compuverde.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Intel® AES-NI Intel®SSD Enterprise storage
- 52. 0 1 2S Intel® Xeon®processor E5-2680 v4 2S Intel® Xeon® Platinum 8168 processor 52 ZebwareZebEC* Application: Zebware’s ZebEC is a novel proprietary erasure code, built for maximum performance on different hardware platforms, reaching architectures from the edge to the cloud. Target customers are storage builders. Potential Customer Benefits: Up to 1800 GB/s throughput for the new platform. Better utilization of hardware investments. Reduced power consumption. Improved redundancy. Portable between hardware platforms. Performance Factors: Overall performance increase of 90%1 was aided by Intel compilers and Intel® AVX-512, which contributed 15%1 of the speed up. “Zebware’s erasure code significant performance boost enables a new generation of Storage solution that have previously not been possible.” – Ramzi Ferchichi, Director of R&D, Zebware Up to 90% faster 1 - Testing conducted on ISV* software comparing 2S Intel® Xeon® Platinum 8168 processor to 2S Intel® Xeon® Processor E5-2680 v4 Testing by Zebware. For complete testing configuration details, see the Configurations Details section, slide 135. www.zebware.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Intel® AVX-512 ATAGLANCE Hardware: Intel® Xeon® 8168 Platinum Processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® C++ Compilers Workload: ZebEC Decoding Throughput (GB/s) 4+2, 1 erasure Zebware ZebEC increased performance with the Intel® Xeon® Platinum 8168 NormalizedPerformance Enterprise
- 53. 0 1 2S Intel® Xeon®processor E5-2697 v3 2S Intel® Xeon® Platinum 8180 processor 53 QuikfyndENTERPRISESEARCH* Application: QuikFynd Enterprise Search allows users to search data in company intranet from multiple data sources. Multiple users query data that is split across 100 databases. For delivering query results, multiple rows from several databases need to be fetched and processed to calculate relevancy rank for the query term. Workload is CPU bound because rank needs to be calculated for all rows that match query criteria. Potential Customer Benefits: More cores in Intel® Xeon® Platinum 8180 processor compared to the previous processor generation allow more queries to be processed in parallel, e.g., more users can be supported on a single server For a given number of users, query response time improves. Performance Factors: More cores and high memory bandwidth of the platform were key to the overall performance gain of 1.79X1 over the previous processor generation. Up to 79% faster QuikFynd* increased performance with the Intel® Xeon® Platinum 8180 processor1 1 - Testing conducted on QuikFynd* software comparing Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2697 v3 Testing done by QuikFynd. For complete testing configuration details, see the Configurations Details section, slide 135 NormalizedPerformance www.quikfynd.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Workload: QuikFynd Enterprise Search* 2.7.4 queries per second databaseEnterprise multi-gen
- 54. 54 InterSystemsCaché* Application: InterSystems’ TrakCare isa unified healthcare information system that enables coordinated care within a hospital or across care settings throughout a region. It creates a consolidated view of each patient’s history, provides secure access to records, enables interoperability with legacy and future applications, and performs analytics that provide real-time insights to drive fully-informed actions. Performance was measured using database accesses per second, known as “GloRefs” Potential Customer Benefits: Encourage hardware refresh for customers currently on Intel® Xeon® E7-4800/8800 v3 series and/or Intel® Xeon® E5-2600 v3 series. Deliver better transaction response times. Demonstrate massive scalability. “TrakCare on Caché has really benefitted from this latest Intel® Xeon Scalable processor, on both 2S and 4S platforms. Although mileage may vary, we would expect all InterSystems Caché based applications to achieve similar improvements.” Mark Bolinsky, Senior Technology Architect ATAGLANCE Hardware: 4S Intel® Xeon® Platinum 8180 processor 2S Intel® Xeon® Platinum 8168 processor Intel® Solid State Drive DC P3700 with with PCIe NVMe 3.0 Interface Platform Features: Full Crossbar interconnect Intel® Ultra Path Interconnect (Intel® UPI) 10.4GT/s 1 - Testing conducted on InterSystems Caché and TrakCare software comparing Intel® Xeon® Platinum 8180 processor to 4S Intel® Xeon® Processor E7-8890 v3 Testing done by Intel. For complete testing configuration details, see the Configurations Details section, slide 135 www.intersystems.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors 0 1 2 4S Intel® Xeon® processor E7-8890 v3 4S Intel® Xeon® Platinum 8180 processor Workload: GloRefs; Transaction Resp Time (seconds) Up to 93% faster InterSystems Caché improved response times with the Intel® Xeon® Platinum 8180 processor1 NormalizedPerformanceNormalizedPerformance 2x Performance InterSystems Caché improved response times with the Intel® Xeon® Platinum 8168 processor1 0 1 2S Intel® Xeon® processor E5-2680 v3 2S Intel® Xeon® Platinum 8168 processor Workload: GloRefs; Transaction Resp Time (seconds) Up to 72% faster 4 socket HLSDATABASEEnterprise multi-gen
- 55. 0 1 2S Intel® Xeon®processor E5-2699 v4 2S Intel® Xeon® Platinum 8180 processor Performance Metric: TPS(Transactions per sec) 55 Altibase7.1* Application: Altibase 7.1* is an in-memory RDMS and provides fast data processing speeds for online transactional processing (OLTP) and online analytical processing (OLAP) workloads. It is aiming for real- time access to time critical data. Potential Customer Benefits: Accelerates time to insights and analytics, due to faster data processing speeds for OLTP and OLAP Greater throughput from improved core scalability Performance Factors: Application benefited from optimizations for Intel® AVX-512 instructions for the Intel® Xeon Scalable processor platform. Highly-parallel processing capability on Intel® Xeon® processors, and Intel® Software Tools also contributed to the boost. Comparison of 4S Intel® Xeon® Platinum 8180 Processor to the 4S Intel® Xeon® processor E7-8890 v4 yielded a 56% performance gain for the same workload. “Altibase 7.1* is aiming for real-time access to time critical data. The Intel® Xeon® Platinum 8180 processor enables Altibase 7.1* to achieve more analytics results and faster transactions storing and manipulating data.” Hess Lee– Senior Managing Director R&D, Altibase ATAGLANCE Hardware: Intel® Xeon® Platinum 8180 processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® C++ Compilers Intel® VTuneTM Up to 75% faster Altibase 7.1* increased performance with the 2S Intel® Xeon® Platinum 8180 processor and Intel® Software Tools1 1 - Testing conducted on Altibase* software comparing 2S Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4 Testing done by Intel. For complete testing configuration details, see the Configurations Details section, slide 135 NormalizedPerformance www.altibase.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable ProcessorsIn-memory Database Intel® AVX-512 Enterprise 4 socket
- 56. 0 1 2S Intel® Xeon®processor E5-2699 v4 2S Intel® Xeon® Platinum 8168 processor Performance Metric: TPS(Transactions per sec) 56 SUNJESOFTGoldilocks* Application: Goldilocks is an in-memory RDBMS with high scalability and low- latency performance. Potential Customer Benefits: Serve more customers, with faster response times. Performance Factors: More cores and 6 memory channels contributed to the performance. Application benefited from optimizations for Intel® AVX-512 instructions, new to Intel® Xeon Scalable processors. Highly-parallel processing capability on Intel® Xeon® processors, and Intel® Software Tools also contributed to the boost. “The 2 socket Intel® Xeon® Platinum 8168 enables Goldilocks to create higher Values with In-Memory Computing, lower TCO, Unlimited Scalability through Scale out Architecture and In Business Continuous Processing area like IIOT, Cloud Computing, Big-Data, Mobile Service” Yongbeom Lee – Executive Director, SUNJESOFT Inc. ATAGLANCE Hardware: Intel® Xeon® Platinum 8168 processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® C++ Compilers Intel® VTune™ Amplifier Up to 61% faster SUNJESOFT GOLDILOCKS* increased performance with the 2S Intel® Xeon® Platinum 8168 processor and Intel® Software Tools1 1 - Testing conducted on SUNJESOFT* software comparing 2S Intel® Xeon® Platinum 8168 processor to 2S Intel® Xeon® Processor E5-2699 v4 Testing done by Intel. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedPerformance www.sunjesoft.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Intel® AVX-512In-memory DatabaseEnterprise
- 57. 0 1 4S Intel® Xeon®processor E7-8880 v4 4S Intel® Xeon® Platinum 8164 processor Performance Metric: TPS(Transactions per sec) 57 SUNJESOFTGoldilocks* Application: Goldilocks is an in-memory RDBMS with high scalability and low- latency performance. Potential Customer Benefits: Serve more customers, with faster response times. Performance Factors: More cores and 6 memory channels contributed to the performance. Application benefited from optimizations for Intel® AVX-512 instructions, new to Intel® Xeon Scalable processors. Highly-parallel processing capability on Intel® Xeon® processors, and Intel® Software Tools also contributed to the boost. “The 4 socket Intel® Xeon® Platinum 8164 enables GoldilocksTM to create higher Values with In-Memory Computing, lower TCO, Unlimited Scalability through Scale out Architecture and In Business Continuous Processing area like IIOT, Cloud Computing, Big-Data, Mobile Service.” Yongbeom Lee – Executive Director, SUNJESOFT Inc. Up to 47% faster SUNJESOFT GOLDILOCKS* increased performance with the 4S Intel® Xeon® Platinum 8164 processor and Intel® Software Tools1 1 - Testing conducted on SUNJESOFT* software comparing 4S Intel® Xeon® Platinum 8164 processor to 4S Intel® Xeon® Processor E7-8880 v4 Testing done by Intel. For complete testing configuration details, see the Configurations Details section, slide 135 NormalizedPerformance www.sunjesoft.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Intel® AVX-512 ATAGLANCE Hardware: Intel® Xeon® Platinum 8164 processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® C++ Compilers Intel® VTuneTM 4 socket In-memory DatabaseEnterprise
- 58. 0 1 2S Intel® Xeon®processor E5-2699 v3 2S Intel® Xeon® processor E5-2699 v4 2S Intel® Xeon® Platinum 8180 processor Up to 25% faster 58 Up to 85% faster 1 - Testing conducted on Innovatrics* software comparing 2S Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v3 and to 2S Intel® Xeon® Processor E5-2699 v4. Testing by Intel and Innovatrics. For complete testing configuration details, see the Configurations Details section, slide 135 NormalizedPerformance www.innovatrics.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors multi-gen Application: Innovatrics Automated Fingerprint Identification System is a large- scale, easy to integrate, standard-compliant and high performance biometric identity management solution for a biometric fingerprint system. AFIS algorithm (1:N) is compares fingerprint templates loaded into memory, and is CPU and memory bound. Potential Customer Benefits: Higher fingerprint matching speed. Higher identification throughput of Automated Fingerprint Identification System. Performance Factors: Multi core scaling and High Memory bandwidth utilization. Vectorization capabilities helped improve performance. “Innovatrics algorithm achieves a speed of 1040 million fingerprint matches per second with the new Intel Xeon Scalable processors. With our unique algorithm, Innovatrics provides the fastest AFIS solution available on the market. Innovatrics AFIS is a fully scalable, multimodal, easy to integrate fingerprint matching solution which continuously demonstrates best in class performance.” Jan Lunter, CEO at Innovatrics Innovatrics*fingerprintmatching ATAGLANCE Hardware: Intel® Xeon® 8180 Platinum processor Platform Features: Multi core High Memory Bandwidth Software Tools/Libraries : Intel® C++ Compiler 2017 OpenMP Innovatrics Fingerprint matching increased performance with the Intel® Xeon® Platinum 8180 processor1 Innovatrics fingerprint identification benchmark, 14 million fingerprints DB loaded in RAM 1.48X Enterprise
- 59. 0 1 4S Intel® Xeon®processor E7-8890 v4 4S Intel® Xeon® Platinum 8180 processor 59 SoftwareagApama* Application: Software AG's Apama Streaming Analytics—supporting predictive analytics—is a global platform for streaming analytics and intelligent automated action on fast-moving big data. The scenario tested is a location aware customer experience management scenario, where retail outlets each want to send a message to a customer if they stray within a certain area around the store. The workload calculates the number of events (location updates) the scenario was capable of processing, within an area of 500K retail outlet geofences. Potential Customer Benefits: Location aware streaming analytics is a key emerging usage segment, and the results from the testing show how choosing Apama 10.0 on the latest Intel® Xeon® Scalable processors significantly improves streaming event processing compared to the performance of the previous processor generation, benefiting Apama customers focused on real time analytics Performance Factors: Apama on 4S Intel® Xeon® Platinum 8180 processor can process 1.48X greater throughput compared to the 4S Intel® Xeon® E7-8890 processor v4. Up to 48% faster Apama increased location filtering throughput with the Intel® Xeon® Platinum 8180 processor1 1 - Testing conducted on Apama software comparing 4S Intel® Xeon® Platinum 8180 processor to 4S Intel® Xeon® Processor E7-8890 v4. Testing by Software AG. For configuration details, see the Configurations Details section, slide 135. www.softwareag.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors 4 socket Workload: Streaming analytics for events (location updates) within an area of 500K retail outlet geofences NormalizedPerformance BigDataanalyticsEnterprise
- 60. 0 1 4S Intel® Xeon®processor E7-8890 v4 4S Intel® Xeon® Platinum 8180 processor System R&D Center of NARI Technology co., LTD 60 NariGridDispatchingSystem(GDS) Application: Nari is a leading solution provider of power and automation technologies in China. Its Realtime DB is the core module of Grid Dispatching System (GDS) and is used to collect, store data from thousands of sensors cross state grid and provides the analysis and dispatching plan function. Potential Customer Benefits: Reduce the response time to provide a better user experience Performance Factors: More cores and high memory bandwidth of the platform were key to the overall performance gain of 1.44X1 over the previous processor generation. “Intel® Xeon® Platinum 8180 demonstrates its great technology advantage. It helps Nari GDS Realtime DB to gain 1.44X performance improvement comparing to previous Xeon platform, and will provide better user experiences to our customers.” Gao Yuan，Technical Expert of NARI Group “Intel® Xeon® Platinum 8180显示了强大的技术优势。Nari GDS 实时数据库 获得了跟上一代平台相 比1.44倍的性能提升，将给我们的客户带来更好的用户体验。” 高原，集团技术专家 Nari Grid Dispatching System (GDS)* increased performance with the Intel® Xeon® Platinum 8180 processor 1 - Testing conducted on NARI* software comparing Intel® Xeon® Platinum 8180 processor to 4S Intel® Xeon® Processor E7-8890 v4 Testing done by Intel. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedPerformance www.sgepri.sgcc.com.cn Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Workload: Grid Dispatching System response time (second) Up to 44% faster 4 socket Enterprise
- 61. 0 1 2S Intel® Xeon®processor E5-2699 v4 2S Intel® Xeon® Platinum 8180 processor Workload: Real-time stock calculation; Benchmark schema derived from TPC-H* queries 61 analytics Application: BOS Handles both OLTP and OLAP workload in sharing “single source of truth”, both master and operational data, on an extremely normalized single database schema. The workload is a mixture of memory bound in fetching records by massive parallel queries and core bound in manipulating the fetched records to construct logical views on demand. Potential Customer Benefits: Deliver sustainable agility in minimizing the impact of DB schema changes while functional requirements keep changing over time. Deliver higher availability in avoiding planned downtime, which was inevitable to release a new application with schema changes. Performance Factors: Higher core count definitely improved responsiveness in practical highly multiplexed workload like real-time stock calculation. “Both higher CPU clock rate and larger CPU cache boosted the performance of core bounded data manipulation step, which raised upper bound of real-time calculation capacity. The more computational power, the less likely to suffer from complexity of cache invalidation.” Yuzo Ishida – Senior Application Engineer, NRI (Architect of BOS) ATAGLANCE Hardware: Intel® Xeon® Platinum 8180 processor Intel® P3710 Data Center Solid State Drive Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Up to 42% faster BOS* increased performance with the Intel® Xeon® Platinum 8180 processor1 1 - Testing conducted on NRI* software comparing 2S Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4 Testing done by NRI. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedPerformance www.nri.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Intel® AVX-512 Nomuraresearchinstitute Businessorientedsolution(BOS)* Enterprise
- 62. 0 1 Gen-to-Gen Intel® Optane™ Technology 4SIntel® Xeon® processor E7 8890 V4 4S Intel® Xeon® Platinum 8180 Processor 4S Intel® Xeon® Platinum 8180 Processor w/ 62 YonyouERP-NC Application: Yonyou ERP-NC is a J2EE-based ERP for large-scale companies, and is widely used in PRC. Potential Customer Benefits: Deliver more TPS to provide a better user experience. Accelerate processing with Intel® Optane™ technology. Performance Factors: More cores and high memory bandwidth of the platform were key to the overall performance gain of 1.42X over previous generation Xeon platform1. High-speed, low-latency access to data with Intel® Optane™ Solid State Drive Data Center P4800X dramatically improves I/O efficiency and delivered a 35% performance improvement1. “Based on Intel® Xeon® Platinum 8180 processor, NC ERP application showed impressively performance boost compared with previous platform. The great performance and reliability fully satisfy NC strict requirements for its mid or large-scale enterprise customers, and it is one of the ideal deployment platforms for NC.” Zhao Yashun, Industry Director “基于Intel® Xeon® Platinum 8180, NC获得了跟上一代平台相比的十分醒目的性能提升，它的高性能和 高可靠性完全满足了NC大中型客户的严格的要求，是我们产品的理想部署平台之一。”赵亚顺, 行业总 监 Yonyou ERP-NC increased performance with the Intel® Xeon® Platinum 8180 processor with Intel® Optane™ SSD 1 - Testing conducted on ISV* software comparing Intel® Xeon® Platinum 8180 processor to 4S Intel® Xeon® Processor E7-8890 v4 Testing done by Intel. For complete testing configuration details, see the Configurations Details section, slide 135 NormalizedPerformance www.yonyou.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Workload: ERP-NC TPS Intel® Optane™ Intel® optane™ Up to 1.42X faster Up to 1.35X faster 4 socket Enterprise
- 63. 0 1 4S Intel® Xeon®processor E7 8890 V4 4S Intel® Xeon® Platinum 8180 Processor 63 KingdomtechnologyKingdomSTS* Application: Kingdom Technology is a leading ISV in financial market industry in China. Kingdom STS transaction middleware is widely deployed for security customers. Potential Customer Benefits: Deliver more transactions per second to provide a better user experience Performance Factors: More cores and high memory bandwidth of the Intel® Xeon® Scalable platform were key to the overall performance gain of 1.4X1 compared to the previous processor generation. “Based on the Intel® Xeon® Platinum 8180 processor, Kingdom STS application showed an impressive performance boost compared to the previous processor generation. The great performance and reliability fully satisfy STS strict requirements for its mid or large-scale security customers, and it is one of the ideal deployment platforms for Kingdom STS.” Zhou Xianqian, Vice GM of Security Software Group, Kingdom “基于Intel® Xeon® Platinum 8180, Kingdom STS获得了跟上一代平台相比的十分醒目的性能提 升，它的高性能和高可靠性完全满足了STS大中型证券客户的严格的要求，是我们产品的理想 部署平台之一。” 周贤谦，金证证券软件总部副总经理 Kingdom STS* increased performance with the 4S Intel® Xeon® Platinum 8180 processor 1 - Testing conducted on Kingdom* software comparing 4S Intel® Xeon® Platinum 8180 processor to 4S Intel® Xeon® Processor E7-8890 v4 Testing done by Intel. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedPerformance www.szkingdom.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable processors Workload: Disabled STS connection to database to measure STS performance Up to 1.4X faster 4 socket Enterprise
- 64. 0 1 2S Intel® Xeon®processor E5-2699 v3 2S Intel® Xeon® processor E5-2699 v4 2S Intel® Xeon® Platinum 8168 processor Up to 1.26X faster 64 QlikQlikView*/QlikSense* Application: BI market leader. Discover deeper insights by building your own rich, guided analytics applications with a flexible, business-ready solution. Workload: internal HWE test scenario, which exercises the CPU by performing calculations on a big data set (1BN rows) Potential Customer Benefits: Deliver a better user experience through faster visualizations. Faster exploration helps generating insights and making better decisions using complex analytics. Could expect even more performance on Intel® Xeon® Platinum 8180 processor. Performance Factors: In-depth and long-term multi-core optimization. Efficient utilization of new enhanced processor cores and cache. “Qlik’s customers place high demands on their compute resources, always pushing for more comprehensive analysis and discovery to gain new insights into business trends and opportunities. That’s why it is crucial to have a high-performance hardware platform with high-availability memory and a powerful processor. Companies that deploy Qlik on the latest Intel® Xeon® Platinum 8168 processor can offer their users a rich, user-friendly visual analytics platform that takes advantage of familiar productivity tools.” Anthony Deighton, CTO & SVP of Products Up to 1.63X faster Qlik Qlikview* increased performance with the Intel® Xeon® Platinum 8168 processor1 1 - Testing conducted on Qlik* software comparing Intel® Xeon® Platinum 8168 processor to 2S Intel® Xeon® Processor E5-2699 v4 and to 2S Intel® Xeon® Processor E5-2699 v4. Testing done by Intel. For complete testing configuration details, see the Configurations Details section, slide 135. NormalizedPerformance www.qlik.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors multi-gen 1.29X AnalyticsEnterprise Workload: QlikView* HWE workload, elapsed time
- 65. 0 1 2S Intel® Xeon®processor E5-2699 v4 2.2GHz/22 cores w/Intel® AVX2 2S Intel® Xeon® Platinum 8180 processor 3.8GHz/28 cores w/Intel® AVX2 65 Oracle Database* 12c Release 2 (12.2), is now available everywhere - in the Cloud, with Oracle Cloud at Customer, and on-premises. This latest release provides organizations of all sizes with access to the world’s fastest, most scalable and reliable database technology in a cost- effective, hybrid Cloud environment. 12.2 also includes a series of innovations that helps customers easily transform to the Cloud while preserving their investments in Oracle Database technologies, skills and resources. The new Intel® Xeon® Platinum 8180 Processor improves Oracle Database 12c Release 2 performance in a 2-socket system by 30%1 compared to the previous generation Intel® Xeon® processor E5-2699 v4. Performance gains measured on the DS workload with both systems utilizing Intel® Advanced Vector Instructions 2 (Intel® AVX2). Improvements is due to higher core count, higher memory bandwidth and new core technology in the latest platform. OracleDatabase*12cRelease2 Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Up to 30% faster NormalizedPerformance Oracle* Database In-Memory Decision Support “DS” Workload 1 - Testing conducted on Oracle* software comparing 2S Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4. Testing by Intel. For configuration details, see the Configurations Details section, slide 135. www.oracle.com Oracle Database improved performance with the Intel® Xeon® Platinum 8180 processor and Intel® AVX-512 databaseEnterprise Intel® Xeon® Scalable Processors
- 66. 0 1 2S Intel® Xeon®Platinum 8180 processor w/Intel® AVX2 2S Intel® Xeon® Platinum 8180 processor w/Intel® AVX-512 66 Oracle Database In-Memory adds in-memory database functionality to transparently accelerate analytic queries by orders of magnitude, enabling real-time business decisions. Using Database In- Memory, businesses can instantaneously run analytics and reports that previously took hours or days. Businesses benefit from better decisions made in real-time, resulting in lower costs, improved productivity, and increased competitiveness. Oracle Database In-Memory accelerates both Data Warehouses and mixed workload OLTP databases, and is easily deployed under any existing application that is compatible with Oracle Database 12c Release 2 (12.2). No application changes are required. Database In-Memory uses Oracle’s mature scale-up, scale-out, and storage-tiering technologies to cost effectively run any size workload. The new Intel® Xeon® Platinum 8180 processor with Intel® Advanced Vector Extensions 512 (Intel® AVX-512) improves Oracle Database In-Memory with Oracle Database 12c Release 2 performance under the DS workload by 20%1 compared to the same system with Intel® Advanced Vector Instructions 2 (Intel® AVX2). OracleDatabaseIn-Memory*withOracleDatabase*12cRelease2 Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Up to 20% faster NormalizedPerformance Oracle* Database In-Memory Decision Support “DS” Workload Intel® AVX-512 1 - Testing conducted on Oracle* software comparing 2S Intel® Xeon® Platinum 8180 processor w/Intel AVX-512 to 2S Intel® Xeon® Platinum 8180 processor w/Intel AVX2. Testing by Intel. For configuration details, see the Configurations Details section, slide 135. www.oracle.com Oracle Database In-Memory* improved performance with the Intel® Xeon® Platinum 8180 processor and Intel® AVX-512 databaseEnterprise Intel® Xeon® Scalable Processors
- 67. 0 1 2S Intel® Xeon®processor E5-2699 v4 2.2GHz/22 cores 2S Intel® Xeon® Platinum 8180 processor 2.7GHz/24 cores 67 Oracle TimesTen In-Memory Database is a memory-optimized relational database that provides applications with extremely fast response time and very high throughput. The new Intel® Xeon® Platinum 8168 processor, with higher core count, higher memory bandwidth and new core technology, improves throughput in a 2-socket system by 19%1 from the previous generation Intel® Xeon® processor E5-2699 v4. OracleTimesTen*In-MemoryDatabase Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Up to 19% faster NormalizedPerformance TimesTen Performance Throughput Benchmark (TPTBM) 100% read 1 - Testing conducted on Oracle* software comparing 2S Intel® Xeon® Platinum 8180 processor to 2S Intel® Xeon® Processor E5-2699 v4. Testing by Intel. For configuration details, see the Configurations Details section, slide 135. www.oracle.com Oracle TimesTen* In-Memory Database improved performance with the Intel® Xeon® Platinum 8180 processor and Intel® AVX-512 databaseEnterprise Intel® Xeon® Scalable Processors
- 68.
- 69. GENERATION- TO-GENERATION 69 Intel®Xeon®ProcessorScalableFamily Software Performance ProofPoints – HPC (Commercial) Potential Customer Benefits: • Faster time to completion for improved TCO • Higher resolution simulations for greater accuracy • More sophisticated analysis for higher confidence in outcomes hpc INTEL® AVX-512 MULTI- GENERATION 1.61XUp to Up to 1.25X Up to5.37X 1 1 – As demonstrated by proof points in this presentation Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others 1 1 Average Gen-to-Gen processor application speedup: 1.4X Average Multi Gen processor application speedup: 1.72X See slide 69 See slide 71 See slide 75
- 70. 0 1 2 2S Intel® Xeon®processor E5-2697 v3 2S Intel® Xeon® processor E5-2699 v4 2S Intel® Xeon® Gold 6148 processor Up to 1.3X faster Up to 2.1X faster 1.61X 70 AltairRADIOSS* Application: RADIOSS is a leading structural analysis solver for highly non-linear problems under dynamic loadings. It is used across all industries worldwide to improve the crashworthiness, safety, and manufacturability of structural designs. Potential Customer Benefits: Run single simulation faster. Run more complex simulations and get more insights by running more simulation options via optimization. Performance Factors: Better performance due to more cores and improved memory bandwidth. Intel® AVX-512 specific executable provides performance improvement with a lower power footprint. Hyper-Threading together with RADIOSS Hybrid MPI OpenMP parallel version deliver an additional 10% performance boost. “The Intel® Xeon® Scalable processor offers amazing computing power for running Altair HyperWorks solvers and especially RADIOSS. Altair is very pleased it becomes now available to our customers.” Eric LEQUINIOU – HPC Director, Altair ATAGLANCE Hardware: Intel® Xeon® 6148 Gold Processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel MPI 2017 Altair RADIOSS 2017 increased performance with the Intel® Xeon® Gold 6148 processor1 1 - Testing conducted on RADIOSS* software comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v3 and to 2S Intel® Xeon® Processor E5-2699 v4. Testing by Altair. For complete testing configuration details, see the Configuration Details section, slide 135. More info See the Solution Brief and Infographic NormalizedPerformance Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors www.altair.com Workload: Summary: NEON front car crash refined model with 1 million of elements multi-genIntel® AVX-512 HPC Manufacturing
- 71. 0 1 2S Intel® Xeon®processor E5-2697 v3 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 1.13X faster Workload: CST EM Studio Time Domain Solver 71 CSTStudioSuite* Application: The electromagnetic simulation software CST STUDIO SUITE® is the culmination of many years of research and development into the most accurate and efficient computational solutions for electromagnetic designs. It comprises CST’s tools for the design and optimization of devices operating in a wide range of frequencies - static to optical. Analyses may include thermal and mechanical effects, as well as circuit simulation. Potential Customer Benefits: Shorter time-to-solution and faster development cycles More accurate and feature rich simulation Better insight into more complex physical problems by more accurate parameter studies Performance Factors: The new Intel® processor has 27% more cores and threads, 50% more memory bandwidth and an improved cache hierarchy. Intel® C++ Compiler, Intel® Math Kernel Library and Intel® MPI Library ATAGLANCE Hardware: Intel® Xeon® 6148 Gold Processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Intel® OmniPath® Architecture Software Tools/Libraries: Intel® C++ Compiler Intel® Math Kernel Library Intel® MPI Library Up to 1.89X faster ISV product* increased performance with the Intel® Xeon® Gold 6148 processor1 1 - Testing conducted on CST* software comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v3 and to 2S Intel® Xeon® Processor E5-2697 v4. Testing by Intel.. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance https://www.cst.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors multi-genIntel® AVX-512 1.47X HPC Manufacturing
- 72. 0 1 2S Intel® Xeon®processor E5-2698 v3 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 1.19X faster 72 Simuliaabaqusstandard Application: Simulia Abaqus Standard gives manufacturers an effective way to analyze static and low-speed dynamic events where precise stress solutions are vital. A single simulation can analyze a model in both the time and frequency domains. Examples include sealing pressure in a gasket joint, steady-state rolling of a tire, or crack propagation in a composite airplane fuselage. Potential Customer Benefits: Faster product design time. Ability to solve more complex models on the same hardware footprint. Performance Factors: Increased core count, higher frequencies and greater memory. bandwidth of the Intel® Xeon® Gold 6148 processor were key to the performance gain. Intel® AVX-512 provides a 25% gain compared to Intel® AVX. ATAGLANCE Hardware: Intel® Xeon® Gold 6148 Processor Intel® Solid State Drive Data Center S3500 Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® Math Kernal Library Intel Fortran Compilers 2016.1 Intel® AVX-512 Intel® SSD Up to 1.8X faster Simulia Abaqus Standard Increased performance with the Intel® Xeon® Gold 6148 processor1 1 - Testing conducted on Simulia* software comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2698 v3 and to 2S Intel® Xeon® Processor E5-2697 v4 Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance https://www.3ds.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors multi-gen Workload: s2b flywheel with centrifugal load “Computational Engineering and Biology are two of the most compute-intensive domains bottlenecking business today. The performance improvements we’ve seen with the Intel® Xeon® Gold 6148 processor can directly translate into better products and reduced time to market for our customers. In the future performance improvements will be measured in number of lives saved.” Steve Levine – Executive Director, The Living Heart Project More info Case Study: The Living Heart Project and Intel® Technologies 1.51X HPC Manufacturing
- 73. 0 1 2S Intel® Xeon®processor E5-2697 v3 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 1.21X faster Workload: Simulia Abaqus Explicit e10 73 Simuliaabaqusexplicit Application: Abaqus Explicit is a finite element analysis product that is well-suited to simulate brief transient dynamic events such as automotive crashworthiness, and ballistic impact. Potential Customer Benefits: Faster product design time. Ability to solve more complex models on the same hardware footprint. Performance Factors: Increased core count, higher frequencies and greater memory bandwidth of the Intel® Xeon® Gold 6148 processor were key to the performance gain. Intel® SSD DC (2x Intel SSD DC S3500 Series RAID0 array). ATAGLANCE Hardware: Intel® Xeon® Gold 6148 Processor Intel® Solid State Drive Data Center S3500 RAID0 array Platform Features: Intel® Advanced Vector Extensions (Intel® AVX) Intel® Hyper-Threading Technology Software Tools/Libraries: Intel Fortran Compilers 2016.1 Intel® SSD Up to 1.74X faster Simulia Abaqus Explicit increased performance with the Intel® Xeon® Gold 6148 processor 1 - Testing conducted on Simulia* software comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v3 and 2S Intel® Xeon® Processor E5-2697 v4 Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors multi-gen “Computational Engineering and Biology are two of the most compute-intensive domains bottlenecking business today. The performance improvements we’ve seen with the Intel® Xeon® Gold 6148 processor can directly translate into better products and reduced time to market for our customers. In the future performance improvements will be measured in number of lives saved.” Steve Levine – Executive Director, The Living Heart Project Intel® AVX https://www.3ds.com More info Case Study: The Living Heart Project and Intel® Technologies 1.43X HPC Manufacturing
- 74. 0 1 2S Intel® Xeon®processor E5-2697 v3 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 1.84X faster 74 LSTCLS-DYNAexplicit* CAE/CrashSimulation Application: LS-DYNA is a popular crash simulation application. It is used by the automobile, aerospace, construction, military, manufacturing, and bioengineering industries in worldwide. Potential Customer Benefits: All major Auto Makers and Aerospace customers can benefit from the increased performance. Faster simulation turnover Influencing customers to migrate to Intel® AVX-512 Platform Features: More cores and threads, 50% more memory bandwidth and an improved cache hierarchy. Additional performance improvement with Intel® AVX-512 ATAGLANCE Hardware: Intel® Xeon® 6148 Gold processor Intel® Solid State Drive Data Center S3520 Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Increased memory bandwidth Wider vector unit Increased number of cores Software Tools/Libraries: Intel® Fortran compiler V16.3 Intel® MPI library 2017 update 1 LS-DYNA explicit increased performance with the Intel® Xeon® Gold 8164 processor1 NormalizedPerformance www.lstc.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Up to 1.25X faster "We are excited to see the potential of the Intel® Xeon® Gold 6148 processor as seen by the node to node comparison leading to faster job turnaround times for our LS-DYNA users. This performance will accelerate crash results in the automotive design phase, reducing time to market.” Nathan Hallquist – Senior Developer LSTCWorkload: 2M elements Car2car model with 120ms simulation time Intel® AVX-512 Intel® SSD 1 - Testing conducted on LS-DYNA* software comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v3 and to 2S Intel® Xeon® Processor E5-2697 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. 1.47X HPC
- 75. 0 1 2S Intel® Xeon®processor E5-2697 v3 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 23% faster Workload: INTES PERMAS non-linear contact simulation of a 19-million nodes engine model. 75 INTESPERMAS*V16 Application: PERMAS by INTES is an advanced Finite Element software system that offers a complete range of physical models at high performance, quality, and reliability. It plays a mission-critical role in the design process at customers from automotive, ship design, aerospace, and more. Potential Customer Benefits: Faster development cycles with shorter time-to-solution for more effective product design. Finer grained physical analysis with larger models at same time- to-solution as with previous generation processors. Deeper insights into complex physics through more sophisticated and more complex analyses. Performance Factors: 1.44X gain from Intel® Xeon® 6148 Gold Processor compared to the Intel® Xeon processor E5-2697 v41. 1.17X gain from Intel® AVX-512 compared to Intel® AVX21. “Our performance optimizations of PERMAS together with the increase of processing power over three generations of Intel processors and Intel NVMe SSDs help our customers reduce their development times and, at the same time, increase the complexity of the analysis required for the design of future products.” Hartmut Manz, Manager PERMAS User Support ATAGLANCE Hardware: Intel® Xeon® 6148 Gold Processor Intel® Solid State Drive DC P3600 with PCIe NVMe 3.0 Interface Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® Composer for Fortran Intel® Math Kernel Library Intel® AVX-512 Intel® SSD Up to 77% faster INTES PERMAS* boosts performance with Intel® Xeon® Gold 6148 processors and Intel® SSD DC P3600 1 - Testing conducted on INTES* PERMAS* comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697v4 and 2S Intel® Xeon® Processor E5-2697v3. Testing by INTES. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance www.intes.de Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors multi-gen 1.44X multi-gen HPC Manufacturing More info See the Solution Brief and Infographic
- 76. 0 1 2 3 4 5 2S Intel® Xeon®processor E5-2680, PERMAS V15 2S Intel® Xeon® processor E5-2697v2, PERMAS V15 2S Intel® Xeon® processor E5-2697v3, PERMAS V16 2S Intel® Xeon® processor E5-2697v4, PERMAS V16 2S Intel® Xeon® Gold 6148 processor, PERMAS V16 Workload: INTES PERMAS non-linear contact simulation of a 19-million nodes engine model. 76 INTESPERMAS*v16 Manufacturing Application: PERMAS by INTES is an advanced Finite Element software system that offers a complete range of physical models at high performance, quality, and reliability. It plays a mission-critical role in the design process at customers from automotive, ship design, aerospace, and more. Potential Customer Benefits: Faster development cycles with shorter time-to-solution for more effective product design Finer grained physical analysis with larger models at same time- to-solution as with previous generation processors Deeper insights into complex physics through more sophisticated and more complex analyses Performance Factors: 1.44X gain from Intel® Xeon® 6148 Gold Processor compared to the Intel® Xeon processor E5-2697 v41 1.17X gain from Intel® AVX-512 compared to Intel® AVX21 “Our performance optimizations of PERMAS together with the increase of processing power over three generations of Intel processors and Intel NVMe SSDs help our customers reduce their development times and, at the same time, increase the complexity of the analysis required for the design of future products.” Hartmut Manz, Manager PERMAS User Support ATAGLANCE Hardware: Intel® Xeon® 6148 Gold Processor Intel® Solid State Drive DC P3600 with PCIe NVMe 3.0 Interface Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® Composer for Fortran Intel® Math Kernel Library Intel® SSD INTES PERMAS* boosts performance with Intel® Xeon® Gold 6148 processors and Intel® SSD DC P3600 1 - Testing conducted on INTES* PERMAS* comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697v4 and 2S Intel® Xeon® Processor E5-2697v3. Testing by INTES. For complete testing configuration details, see the Configuration Details section, slide 135. More info See the Solution Brief and Infographic NormalizedPerformance www.intes.de Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors multi-gen Up to 5.37X Intel® AVX-512
- 77. 0 1 4S Intel® Xeon®processor E7-8890 v4, 256GB 4S Intel® Xeon® processor E7-8890 v4, 4TB 4S Intel® Xeon® Platinum 8180 processor Up to 56% faster Workload: INTES PERMAS non-linear contact simulation of a 19-million nodes engine model. 77 INTESPERMAS*V16 Application: PERMAS by INTES is an advanced Finite Element software system that offers a complete range of physical models at high performance, quality, and reliability. It plays a mission-critical role in the design process at customers from automotive, ship design, aerospace, and more. Potential Customer Benefits: Faster development cycles with shorter time-to-solution for more effective product design. Finer grained physical analysis with larger models at same time-to- solution as with previous generation processors. Deeper insights into complex physics through more sophisticated and more complex analyses. Performance Factors: Compelling price/perf ratio for the Intel® SSD DC P3600 with PCIe NVMe compared to DRAM on the Intel® Xeon® processor E7-8890 v4. 1.42X gain compared to the Intel® Xeon® processor E7-8890 v41. > 4 TFLOP sustained compute performance in numerical kernels. “Our performance optimizations of PERMAS together with the increase of processing power over three generations of Intel processors and Intel NVMe SSDs help our customers reduce their development times and, at the same time, increase the complexity of the analysis required for the design of future products.” Hartmut Manz, Manager PERMAS User Support ATAGLANCE Hardware: Intel® Xeon® 6148 Gold Processor Intel® Solid State Drive DC P3600 with PCIe NVMe 3.0 Interface Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® Composer for Fortran Intel® Math Kernel Library Intel® SSD INTES PERMAS* boosts performance with Intel® Xeon® Gold 6148 processors and Intel® SSD DC P3600 1 - Testing conducted on INTES* PERMAS* comparing Intel® Xeon® Platinum 8180 processor to 4S Intel® Xeon® Processor E7-8890v4. Testing by INTES. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance www.intes.de Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors 1.42x Intel® AVX-5124 socket HPC Manufacturing More info See the Solution Brief and Infographic
- 78. 0 1 2 3 4S Intel® Xeon®processor E7-8890v3, PERMAS V15 4S Intel® Xeon® processor E7-8890v4, PERMAS V15 4S Intel® Xeon® processor E7-8890v4, PERMAS V16 4S Intel® Xeon® processor Platinum 8180, PERMAS V16 Workload: INTES PERMAS non-linear contact simulation of a 19-million nodes engine model. 78 INTESPERMAS*v16 Application: PERMAS by INTES is an advanced Finite Element software system that offers a complete range of physical models at high performance, quality, and reliability. It plays a mission-critical role in the design process at customers from automotive, ship design, aerospace, and more. Potential Customer Benefits: Faster development cycles with shorter time-to-solution for more effective product design Finer grained physical analysis with larger models at same time-to- solution as with previous generation processors Deeper insights into complex physics through more sophisticated and more complex analyses Performance Factors: Compelling price/perf ratio for the Intel® SSD DC P3600 with PCIe NVMe compared to DRAM on the Intel® Xeon® processor E7-8890 v4. 1.42X gain compared to the Intel® Xeon® processor E7-8890 v41. > 4 TFLOP sustained compute performance in numerical kernels. “Our performance optimizations of PERMAS together with the increase of processing power over three generations of Intel processors and Intel NVMe SSDs help our customers reduce their development times and, at the same time, increase the complexity of the analysis required for the design of future products.” Hartmut Manz, Manager PERMAS User Support ATAGLANCE Hardware: Intel® Xeon® 6148 Gold Processor Intel® Solid State Drive DC P3600 with PCIe NVMe 3.0 Interface Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® Composer for Fortran Intel® Math Kernel Library INTES PERMAS* boosts performance with Intel® Xeon® Gold 6148 processors and Intel® SSD DC P3600 1 - Testing conducted on INTES* PERMAS* comparing Intel® Xeon® Platinum 8180 processor to 4S Intel® Xeon® Processor E7-8890v 4 and 4S Intel® Xeon® Processor E7-8890 v3. Testing done by INTES. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance www.intes.de Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors multi-gen 2.97x INTESPERMAS*V16 Intel® SSDIntel® AVX-512 4 socket HPC Manufacturing More info See the Solution Brief and Infographic
- 79. 0 1 2S Intel® Xeon®processor E5-2680 v3 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor 79 MSCNastran* Application: MSC Nastran is a multidisciplinary structural analysis application used by engineers to perform static, dynamic, and thermal analysis across the linear and nonlinear domains, complemented with automated structural optimization and award winning embedded fatigue analysis technologies, all enabled by high performance computing. Potential Customer Benefits: Faster simulation turnover. Better designs, improved products. Performance Factors: Intel® AVX-512. Optimized Intel® MKL. Better memory bandwidth (up to 15%). Better I/O bandwidth (up to12%). Among three workloads, the geomean performance improvement for the new Intel® Xeon® Gold 6148 processor is 1.35X compared to the previous processor generation. “MSC Nastran heavily relies on Intel MKL technology in Lanczos and ACMS and thus can make full use of the advances in the Intel Xeon Gold processor” – Dr. Travis Austin, Development Director, MSC Nastran ATAGLANCE Hardware: Intel® Xeon® 6148 Gold processor Intel® Solid State Drive Data Center P3700 Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Increased memory bandwidth Software Tools/Libraries: Intel® Fortran compiler Intel® Math Kernal Library Intel® MPI library 2017 update 1 Up to 74% faster MSC Nastran increased performance with the Intel® Xeon® Gold 6148 processor1 Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others CAE/FEASimulation Up to 33% faster Intel®SSDIntel® AVX-512 Workload: xx1cmd0 SOL103/w Lanzcos model 10 cores used Intel® Xeon® Scalable Processors 1.3X 1 - Testing conducted on MSC* software comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2680 v3 and to 2S Intel® Xeon® Processor E5-2697 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformanceHPC www.mscsoftware.com
- 80. 0 1 2S Intel® Xeon®processor 2697 v3 2S Intel® Xeon® processor 2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 32% faster Workload: Lemans 17M 80 SIEMENSPLMSTAR-CCM+* Application: STAR-CCM+ is a Computer Aided Engineering (CAE) application that provides a complete multidisciplinary platform for the simulation of products and designs operating under real-world conditions. Solving complex industrial problems requires a simulation tool that spans a multitude of physical phenomena and a variety of engineering disciplines. Potential Customer Benefits: More cores per node enables faster simulations. Ideal for compute clusters. Performance Factors: Benchmarks completed with Intel® MPI 2017 U2. “Intel® Xeon® Gold 6148 processors are certified with v12.04, the very latest version of STAR-CCM+. This enables our customers to get quicker answers for their engineering simulations and achieve better designs, faster.” Keith Foston - HPC Product Manager ATAGLANCE Hardware: Intel® Xeon® Gold 6148 processor Platform Features: More cores Better memory bandwidth Software Tools/Libraries Intel® MPI 2017 Up to 70% faster Siemens STAR-CCM+* 12 increased performance with the Intel® Xeon® Gold 6148 processor 1 - Testing conducted on Siemens* software comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v3 and to 2S Intel® Xeon® Processor E5-2697 v4. Testing done by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® scalable Processors multi-gen manufacturing https://mdx.plm.automation.siemens.com/star-ccm-plus HPC 1.28X
- 81. 81 Application: ANSYS® Fluent isa general purpose CFD and multiphysics solver widely used in automotive manufacturing, aerospace, academia, and Formula 1 racing. Typical workload sizes range from 2 million to 500 million cells. Fluent scalability has been demonstrated to 1000’s of Intel® Xeon® processors with Intel® Omni-Path Fabric. Potential Customer Impact: Increase single node performance, smaller cluster footprint or higher performance for the same footprint. Per-core performance improvement, based on a 32 core run on each platform for apples-to-apples license utilization: Intel® Xeon® Gold 6148 processor outperforms the Intel® Xeon® processor E5-2697 v4 by 34% on sedan_4m Intel® Xeon® Gold 6148 platform outperforms the Intel® Xeon® processor E5-2698 v3 by 46% on sedan_4m Performance Factors: Intel® AVX2 build provides up to 12% improved performance, primarily of benefit for polyhedral cell types. Better per-core performance due to improved memory bandwidth. ATAGLANCE Hardware: Intel® Xeon® 6148 Gold Processor Platform Features: More cores More memory bandwidth Software Tools/Libraries Compiler/MPI: Intel® Parallel Studio XE Cluster Edition Up to YY% faster www.ansys.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors multi-gen “ANSYS teamed with Intel to make sure software and hardware improvements go hand in hand. The latest combination of ANSYS Fluent 18.1 and Intel® Xeon® Gold 6148 is a clearly testament of impressive overall performance gains achieved for customers who want to increase their engineering productivity.” Dr. Wim Slagter – Director of HPC and cloud marketing, ANSYS 0 1 2S Intel® Xeon® processor E5-2698 v3 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 13% faster Fluent workload: sedan_4m. ANSYS® Fluent 18.1 increased performance1 with the Intel® Xeon® Gold 6148 processor Up to 60% faster Ansys®fluent 1.41X 1 - Testing conducted on ANSYS* software comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2698 v3 and to 2S Intel® Xeon® Processor E5-2697 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. HPC Manufacturing More info See the infographic
- 82. 0 1 2S Intel® Xeon®processor E5-2698 v3 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 12% faster ANASYS Mechanical workload: V18cg-3 ANSYS® Mechanical 18.1 increased performance1 with the Intel® Xeon® Gold 6148 processor Up to 59% faster 82 ANSYS®Mechanical HPC Application: ANSYS® Mechanical* is a structural analysis and multi-modal solver widely used in manufacturing industries for both linear and non- linear static and transient analysis. Typical workload sizes range from a few million degrees of freedom and up. ANSYS® Mechanical leverages Intel® MKL to support the latest Intel Instruction Set Architecture. Potential Customer Benefits: Increase performance, faster analysis, more design iterations possible. Per-core performance improvement, based on a 32 core run on each platform for apples-to-apples license utilization: Intel® Xeon® Gold 6148 processor outperforms the Intel® Xeon® processor E5-2697 v4 by 36% on V18cg-3 Intel® Xeon® Gold 6148 processor outperforms the Intel® Xeon® processor E5-2698 v3 by 51% on V18cg-3 Performance Factors: Better per core performance and better scaling due to improved memory bandwidth. Higher core count available for solver faster turnover. ATAGLANCE Hardware: Intel® Xeon® 6148 Gold Processor Platform Features: More cores More memory bandwidth Software Tools/Libraries Compiler/MPI: Intel® Parallel Studio XE Cluster Edition Intel® Math Kernel Library (Intel® MKL) www.ansys.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors multi-gen “Thanks to the latest Intel® Xeon® Gold 6148, ANSYS Mechanical 18.1 can take full advantage of higher core counts and more on-board memory so that users can consider more product ideas, easily making design tradeoffs and increasing their engineering productivity.” Dr. Wim Slagter – Director of HPC and cloud alliances, ANSYS 1 - Testing conducted on ANSYS* software comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2698 v3 and to 2S Intel® Xeon® Processor E5-2697 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. 1.41X Manufacturing
- 83. 0 1 2S Intel® Xeon®processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 50% faster 83 ANSYSCFX* Application: ANSYS CFX is a high-performance computational fluid dynamics (CFD) software tool that delivers reliable and accurate solutions quickly and robustly across a wide range of CFD and multi-physics applications. This application is memory bandwidth and latency bound. Potential Customer Benefits: Faster time to market. Solve larger tasks with the same total cost of ownership. Performance Factors: By taking advantage of Intel® Xeon® Gold 6148, ANSYS CFX experienced up to 50% performance gain compared to the previous processor generation. Improved memory system delivers results faster. Intel® AVX-512 contributes up to a 10% speedup. Ansys CFX v18* increased performance with the Intel® Xeon® Gold 6148 processor1 1 - Testing conducted on CFX* software comparing 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v4 Testing by Intel and ANSYS. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance www.ansys.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors ATAGLANCE Hardware: Intel® Xeon® 6148 Gold Processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) More cores More memory bandwidth Software Tools/Libraries Fortran Compiler/MPI: Intel® Parallel Studio XE Cluster Edition Intel® AVX-512 Workload: perf_Airfoil_10M_R14 HPC Manufacturing
- 84. 0 1 2S Intel® Xeon®processor E5-2699 v4 2S Intel® Xeon® Platinum 8168 processor 84 IHSMARKITMarkitAnalyticsriskengine* Application: Markit Analytics provides a range of enterprise risk management software solutions, using the latest risk technology to deliver exceptional computation speed and rapid time to market. This application is core/memory bound. Potential Customer Benefits: Faster turnaround and time-to-market. Solving larger tasks with a comparable TCO. Performance Factors: Intel compilers and Intel® AVX-512 delivered vectorized code, contributing to the improved performance. ATAGLANCE Hardware: Intel® Xeon® Platinum 8168 processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® Parallel Studio XE 2017 Up to 58% faster ISV product* increased performance with the Intel® Xeon® Platinum 81681 1 - Testing conducted on IHS Markit* software comparing Intel® Xeon® Platinum 8168 processor to 2S Intel® Xeon® Processor E5-2699 v4. Testing by IHS Markit and Intel. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance www.ihsmarkit.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® scalable Processors Intel®AVX-2FinancialservicesindustryHPC Workload: IHS Markit Analytics Risk Engine internal synthetic portfolio.
- 85. 0 1 2S Intel® Xeon®processor E5-2680 v3 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor 85 NXNastran Application: NX Nastran is a finite element (FE) solver for stress, vibration, buckling, structural failure, heat transfer, acoustics and aeroelasticity analyses. Manufacturers as well as engineering suppliers in aerospace, automotive, electronics, heavy machinery, medical device, and other industries rely on NX Nastran software for their critical engineering computing needs so they can produce safe, reliable and optimized designs within increasingly shorter design cycles. Potential Customer Benefits: Faster simulation turnover. Better designs, improved products. Performance Factors: Intel® AVX-512. Optimized Intel® MKL. Improved memory bandwidth. Improved I/O bandwidth. “Intel Xeon Scalable Processors provides clear performance benefits on very demanding NX Nastran simulation workflows. AVX-512 in particular is a good fit for many high-fidelity engineering simulations.” Robert Crockett, Software Engineering Manager, Siemens Dynamics and Numerical Methods Group Siemens NX Nastran increased performance with the Intel® Xeon® Gold 6148 processor1 NormalizedPerformance www.siemens.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Up to 17% faster Up to 52% faster Intel®SSDIntel® AVX-512 Unrestricted Workload: Input_deck model: 3.5M DOFs simplified automotive body-in-white model for Modal Frequency Response analysis. 1 - Testing conducted on NX NASTRAN* software comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2680 v3 and to 2S Intel® Xeon® Processor E5-2697 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. ATAGLANCE Hardware: Intel® Xeon® 6148 Gold processor Intel® Solid State Drive Data Center P3700 Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Increased memory bandwidth Software Tools/Libraries: Intel® Fortran compiler Intel® Math Kernal Library Intel® MPI library 2017 update 1 Intel® Xeon® Scalable Processors 1.29X CAE/FEASimulationHPC More info See the Solution Brief and Infographic
- 86. 0 1 2S Intel® Xeon®processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Workload: PAM-CRASH elapsed seconds neon_front 1M 60ms 86 ESI-GroupVPS* Application: Virtual Performance Software (VPS) originated from a crash simulation software, PAM-CRASH, and is used by manufacturers in order to virtually validate their new designs. The software is based on finite element methods (FEM) and enables the modeling of complex geometry by offering different structural and continuum elements: beams, shells, membranes and solids. Potential Customer Benefits: Shorter time-to-solution for effective design validation. Better performance of VPS will be reflected in shorter simulation time for ESI-Group customers. For example, accelerated crash results in the automotive design phase reduces time to market.2 Performance Factors: The new Intel® processor has 27% more cores and threads, 50% more memory bandwidth and an improved cache hierarchy. Up to 1.43X speedup over Intel® Xeon® processor E5- 2697v4 ATAGLANCE Hardware: Intel® Xeon® Gold 6148 Processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® Fortran Compiler, Intel® MPI Library Up to 1.43X faster ESI VPS increased performance with the Intel® Xeon® Gold 6148 processor1 NormalizedPerformance www.esi-group.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® gold Processor Family 1 - Testing conducted on ESI Group* VPS* software comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v4. Testing done by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. 2 - This proof point is limited to pure crash model. Safety is not included in these benchmarks Intel® AVX-512 HPC Manufacturing
- 87. 0 1 PxrPathTracer.rib PxrUPBP.rib 2S Intel®Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 Processor 87 Pixar’sRENDERMAN21 digital contentcreation Application: Pixar’s RenderMan is the industry standard rendering solution for high-end visual effects and feature animation. It is constantly evolving and the current RIS ray tracing framework is a state-of-the-art system optimized for physically-based rendering effects. The test workloads are example scene files bundled with official RenderMan 21 release. PxrPathTracer.rib uses the traditional path tracing algorithm to render a scene with geometric light object lighting objects with reflective and refractive materials. PxrUPBP.rib uses an advanced ray tracing integrator to compute a realistic simulation of light passing through a dielectric object inside a volumetric environment casting caustics in the volume and other objects. Potential Customer Benefits: Increased performance allows artists to create more realistic scenes and special effects. Faster turn around time for more iteration to achieve their artistic vision. Performance Factors: Up to 37% gain1 from improved core performance and increased core counts. ATAGLANCE Hardware: Intel® Xeon® Gold 6148 Processor Platform Features: More cores and threads, 50% more memory bandwidth and an improved cache hierarchy. Software Tools/Libraries: Intel® Parallel Studio XE 2017 Composer Edition (C++) 15.0 1 - Testing conducted on Pixar’s RenderMan 21 software comparing 2S Intel® Xeon® Processor E5-2697 v4 to 2S Intel® Xeon® Gold 6148 processor. Testing done by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. https://renderman.pixar.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Up to 1.37X faster Up to 1.3X faster Pixar’s RenderMan 21 increased performance with the 2S Intel® Xeon® Gold 6148 processor NormalizedPerformance Workload: PxrPathTracer.rib, PxrUPBP.rib (both bundled with commercial released version) HPC
- 88. 0 1 2S Intel® Xeon®processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Performance Metric: Total execution time Up to 21% faster Up to 30% faster 88 Gaussian* quantum chemistry Application: Gaussian* is a widely used computational chemistry package for electronic structure modeling that encompasses support for semi-empirical, density functional theory (DFT), and ab initio molecular orbital calculations. Depending on the model chemistry (i.e., combination of method and basis set) the workload exhibits diverse characteristics. Apinehfb and apinefreqb are two benchmarks that correspond to the energy and frequency calculation of alpha-pinine, respectively. Hismp4 encompasses the MP4 method for histidine. Potential Customer Benefits: Enables scientists to achieve faster execution of workloads of interest, solve similar problems with higher accuracy, or solve new/larger/more complex problems. Performance Factors: The results were obtained using the official/certified AVX2 binary supplied by Gaussian. This application is enabled with Intel® AVX2 and OpenMP parallelization. ATAGLANCE Hardware: Intel® Xeon® Gold 6148 processor Platform Features: Intel® Advanced Vector Extensions 2 (Intel® AVX2) More cores and threads, 50% more memory bandwidth and an improved cache hierarchy. Up to 22% faster Increased Gaussian* performance with the Intel® Xeon® Gold 6148 processor1 1 - Testing conducted on Gaussian* software comparing 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance www.gaussian.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® gold Processor Family apinehfb apinefreqb hismp4 Intel® AVX2 HPC
- 89. 0 1 2S Intel® Xeon®processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Workload: Parallel Ocean Program stitching latency 89 FIOPOP Application : POP (Parallel Ocean Program) is an ocean circulation model widely used in ocean and climate research. It is also incorporated into FIO-ESM v1.0 (First Institute of Oceanography-Earth System Model version 1.0) as the ocean component. Potential Customer Benefits: Improved performance delivers faster computations & simulations. Boost ocean science development. Performance Factors: More cores and threads, 50% more memory bandwidth and an improved cache hierarchy of the new platform contributed to the performance increase. “High-resolution POP has very high requirement for platform performance. Great to see good performance improvement on the new Intel® Xeon® Gold 6148 processor. This will fasten the simulation process of our application, and it will be one of our preferred platforms after further test.” Haixing Liu, Director of HPC Center, FIO, SOA “高分辨率POP对平台性能有着非常高的要求。在Intel® Xeon® Gold 6148 processor上，我们看到了很好的性能提升，这将加速我们应用的模拟进程， 进一步测试后将是我们应用的理想平台之一。”，刘海行，国家海洋局第一海 洋研究所高性能计算中心主任 ATAGLANCE Hardware: Intel® Xeon® 6148 Platinum Processor Software Tools/Libraries: Intel® C++ Compiler Up to 39% faster FIO POP* increased performance by 1.39X with the Intel® Xeon® Gold 6148 processor1 1 - Testing conducted on POP* software comparing 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v4 Testing done by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance http://www.fio.org.cn/ Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable ProcessorsHPC weather
- 90. 0 1 2S Intel® Xeon®Processor E5-2699 v4 2S Intel® Xeon® Platinum 8180 Processor 31% faster 33% faster 90 Intelembree Application: Path Tracer Renderer using Embree. The code represents a typical ray tracing rendering pipeline used throughout DCC to show comparative performance on different types of hardware with a variety of input 3D data models. Potential Customer Impact: Ability to achieve competitive performance and the flexibility of IA for rendering and render farm applications. Embree and Software Defined Visualization (SDVis) are software image rendering alternatives to processing with GPUs. Performance Factors: Embree on dual socket (56 cores total) Intel® Xeon® Platinum 8180 Processor with Intel® AVX-512 performs faster than dual socket (44 cores total) Intel® Xeon® Processor E5-2699 v4. ATAGLANCE IA + Adjacencies: Intel® Xeon® Platinum 8180 Processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) More cores and threads, 50% more memory bandwidth and an improved cache hierarchy. 31% faster Increased Embree performance with the Intel® Xeon® Platinum 8180 Processor1 1 - Testing conducted on Embree 2.16.1 software comparing 2S Intel® Xeon® Platinum 8180 Processor to 2S Intel® Xeon® Processor E5-2699 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance http://embree.github.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Platinum Processor Family Power Plant 12.8M Tris Crown 4.8M Tris Bentley 2.3M Tris Workload: 1024x1024 image resolution. Frames Per Second Intel® AVX-512 More info Software Defined Visualization digital contentcreationHPC
- 91. 0 0.5 1 2S Intel® Xeon®processor E5-2699 v4 2S Intel® Xeon® Platinum 8168 processor 91 ItivitiTbricks2.10* Application: Itiviti Tbricks’ Smart Order Router (SOR) accepts orders from clients and distributes them among multiple exchanges to get the best execution without moving the market significantly and preventing arbitrage of the client orders between exchanges. Potential Customer Benefits: Lower latency of the SOR (16% lower latency in 90% cases) directly impacts the execution quality and reduces slippage (difference between desired execution price and price available when order reaches the market). Handling bursts efficiently is of exceptional importance for the business – customers trust the ability of SOR to execute efficiently even during the volatile markets. Performance Factors: Improved memory cache hierarchy microarchitecture of the Intel® Xeon® Scalable processor platform contributed to lower latency. ”With the increased core count and new generation architecture we have observed more consistent latency during activity bursts and a significant latency improvement (up to 40%) for some specific workloads (e.g. parallel processing of multiple orders in the Order Entry Market Gateway component).” – Steve Christinson, CTO of Itiviti ATAGLANCE Hardware: Intel® Xeon® 8168 Platinum processor Intel® Solid State Drive Data Center P3520 Platform Features: Rebalanced Cache Hierarchy Latency tuning Software Tools/Libraries: Intel® VTune™ AmplifierXE Up to 40% lower latency Itivity Tbricks* increased performance with the Intel® Xeon® Platinum 8168 processor and the Intel® SSD DC 3700 1 - Testing conducted on Itivity* software comparing Intel® Xeon® Platinum 8168 processor to 2S Intel® Xeon® Processor E5-2699 v4. Testing by Itivity.. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance–LOWERISBETTER www.itiviti.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® scalable Processors Intel®SSD Workload: Smart Order Router in a highly volatile market simulation, measuring latency. LOWER IS BETTER FinancialservicesindustryHPC
- 92. 0 1 2S Intel® Xeon®processor 2699 v4 2S Intel® Xeon® Platinum 8168 processor Workload: Samples trained / sec 92 GenerativeModelscharactergeneration* Application: This trains a generative model to identify and learn patterns within text. After training, the model generates text that resembles the text in the training data. This uses a neural network model called the LSTM Potential Customer Benefits: Faster training performance on the latest Intel® Xeon® Scalable platforms. Faster time to prediction in a field where every millisecond counts. Performance Factors: Used Intel® Math Kernal Library for additional performance gains. Up to 38% faster 1 - Testing conducted on Julia Computing* software comparing 2SIntel® Xeon® Platinum 8168 processor to 2S Intel® Xeon® Processor E5-2699 v4 Testing by Julia Computing. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others MachineLearning Intel® Xeon® Scalable processorsHPC ATAGLANCE Hardware: Intel® Xeon® 8168 Platinum processor Platform Features: More cores More memory bandwidth Intel® Mesh Architecture Software Tools/Libraries: Intel® Math Kernal Library Julia Computing improved performance with the Intel® Xeon® Platinum 8168 processor www.juliacomputing.com
- 93. 0 1 2S Intel® Xeon®processor 2699 v4 2S Intel® Xeon® Platinum 8168 processor Workload: Training performance 93 StockPriceEstimationlinearregression* MachineLearning Application: Linear Regression is used to identify trends and fluctuations in stock prices and time series analysis. This used a multi-layer perceptron to identify stock price patterns. Potential Customer Benefits: Faster training performance on the latest Intel® Xeon® Scalable platforms. Faster time to prediction in a field where every millisecond counts. Performance Factors: Used Intel® Math Kernal Library for additional performance gains. ATAGLANCE Hardware: Intel® Xeon® 8168 Platinum processor Platform Features: More cores More memory bandwidth Intel® Mesh Architecture Software Tools/Libraries: Intel® Math Kernal Library Up to 34% faster Julia Computing improved performance with the Intel® Xeon® Platinum 8168 processor NormalizedPerformance www.juliacomputing.com Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable processorsHPC 1 - Testing conducted on Julia Computing* software comparing 2SIntel® Xeon® Platinum 8168 processor to 2S Intel® Xeon® Processor E5-2699 v4 Testing by Julia Computing. For complete testing configuration details, see the Configuration Details section, slide 135.
- 94. 0 1 2S Intel® Xeon®processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Workload: Actran Acoustic Finite Element calculation – Unsymmetrical complex matrix factorization 94 FreeFieldTechnologiesActran* Application: Actran is a simulation software used by industrials from various industries (automotive, aeronautics, consumer goods…) to improve acoustics, vibro-acoustics and aero-acoustics performances of their products. Potential Customer Benefits: Get simulation results faster. Facilitate usage of optimization algorithms. Performance Factors: By taking advantage of more cores and memory with the Intel® Xeon® Gold 6148 processor, Free Field Technologies experienced a 35% gain1 compared to the previous processor generation. The use of Intel® MKL PARDISO solver allows to factorize efficiently large unsymmetrical complex matrix. The 2666MHz RAM embedded in Intel® Xeon® Gold 6148 improves significantly the solver performance. ATAGLANCE Hardware: Intel® Xeon® Gold 6148 processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® Math Kernal Library Up to 35% faster Free Field Technologies Actran increased performance with the Intel® Xeon® Gold 6148 1 - Testing conducted on Free Field Technologies’ Actran software comparing 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® E5-2697 v4 processor. Testing done by Free Field Technologies. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance www.fft.be Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Intel® AVX-512 CAE/FEASimulationHPC
- 95. 0 1 2S Intel® Xeon®processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 1.21x faster Workload: Flame-D 95 CONVERGECFD* Application: Converge CFD is a computational fluid dynamics software which provides solutions for complex transient problems with moving geometries. It has been used for various applications such as internal combustion engines, gas turbines, compressors. Potential Customer Benefits: Faster product design time Ability to solve more complex models on the same hardware footprint Performance Factors: Increased core count, higher frequencies and greater memory bandwidth of the Intel® Xeon® Platinum 8168 processor were key to the performance gain. Intel® Solid State Drive Data Center S3500 RAID0 array and Intel® AVX-512 provide a performance boost. ATAGLANCE Hardware: Intel® Xeon® Gold 6148 Processor Intel® Solid State Drive Data Center S3500 RAID0 array Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® Fortran Compilers 2017.0 Intel® SSD Up to 1.29X faster CONVERGE CFD increased performance with the Intel® Xeon® Gold 6148 processor1 1 - Testing conducted on Converge* software comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v4 Testing done by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance https://convergecfd.com/ Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Intel® AVX-512 HPC Manufacturing
- 96. 1.00 2.00 4.00 8.28 17.78 1.22 2.67 5.33 10.67 22.86 0 5 10 15 20 1 node 2node 4 node 8 node 16 node 2S Intel® Xeon® E5-2697 v4 processor 2S Intel® Xeon® Gold 6148 processor Workload: NASA CRM with 5 Million Finite Volumes. 96 cfd ATAGLANCE Hardware: Intel® Xeon® Gold 6148 processor Intel® Solid State Drive Data Center Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Intel® Omni-Path Architecture More cores Increased memory bandwidth Software Tools/Libraries: Intel® Fortran Compiler Intel® MPI 1 - Testing conducted on HiFUN software comparing 2S Intel® Xeon® Gold 6148 to 2S Intel® Xeon® E5-2697 v4 processor. Testing done Intel. For complete testing configuration details, see the Configuration Details section, slide 135. Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Intel® AVX-512 Up to 28% faster Application: The flow solver HiFUN (High Resolution Flow Solver on Unstructured Meshes) is a general purpose tool based on unstructured data based cell centre finite volume formulation. HiFUN solves flow problems encountered in aerospace and automotive industries. Potential Customer Benefits: Improved time to solution and improved multi-node performance can lead to increased productivity for HiFUN users. Performance Factors: By taking advantage of Intel® AVX-512 and Intel® Omni-Path Architecture, developers experienced a performance enhancement of 33% at 4 nodes and 28% for 16 nodes compared to the previous processor generation. These results exceeded customers’ expectations. S&IHIFUNSOLVERS & I Engineering Solutions Pvt. Ltd. A Product of Research http://www.sandi.co.in S & I HiFUN increased performance with the Intel® Xeon® Gold 6148 processor1 NASACRM/timeperiteration HPC
- 97. 0 1 1 node 4node 16 node 2S Intel® Xeon® E5-2697 v4 processor 2S Intel® Xeon® Gold 6148 processor Workload: 1025x513x1025 problem size. Upto40%faster 97 Application: DNS-TBL (Direct Numerical Simulation of a spatially developing Turbulent Boundary Layer) is implicit velocity decoupling solver for the incompressible Navier-Stokes equation developed by KAIST and KISTI. Potential Customer Benefits: Faster analysis and research cycles on an Intel® Xeon® Gold 6148 processor cluster with Intel® Omni-Path Architecture. Run larger workloads with an increased number of time steps. Performance Factors: Up to 40% performance improvement with 1 node Intel® Xeon® Gold 6148 processor compared to the previous processor generation. By taking advantage of Intel® AVX-512 and Intel® Omni-Path Architecture, developers experienced a performance increase of up to 28% at 16 nodes. ATAGLANCE Hardware: Intel® Xeon® Gold 6148 processor Platform Features: Intel® Omni-Path Architecture Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® Fortran Compiler Intel® MPI Intel® Math Kernel Library DNS-TBL* increased performance1 with 2S Intel® Xeon® Gold 6148 processor 1 - Testing conducted on DNS-TBL software comparing 2S Intel® Xeon® Gold 6148 processor with Intel® Omni-Path Architecture to 2S Intel® Xeon® Processor E5-2697 v. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® AVX-512Intel®opa www.kaist.ac.kr/html/en www.kisti.re.kr/eng Kaist/kistiDNS-TBL* Intel® Xeon® Scalable Processors Upto28%faster SimulationHPC
- 98. 0 1 2S Intel® Xeon®Gold 6148 processor w/Intel® SSD DC P3700 2S Intel® Xeon® Gold 6148 processor w/Intel® Optane™ SSD Workload: XDU MOM-LASPCEM workload, run time (second) 98 XidianUniversityMoM-LASPCEM Application: MOM-LASPCEM is an in-house large scale parallel MoM code developed by XDU (Xidian University). MoM (method of moments) is a numerically accurate method in computational electromagnetics. By using parallel computing technology, complex electromagnetic problems, such as antenna design for mobile communications efficiently, can be simulated. Potential Customer Benefits: Accelerate MoM code performance, by up to 26%1, and reduce computational time with Intel® Xeon® Scalable processors and with Intel® Optane™ SSDs compared to the previous generation SSD. Users can run simulations faster, produce improved products, and get them to market faster. Performance Factors: Intel® Optane™ SSDs. “26% performance improvement from Intel Optane SSD means a lot to our HPC applications, and will greatly helps us to fasten the research process!” Zhao Xunwang, Vice Director of Shaanxi Key Laboratory of Very Large Scale Electromagnetic Computing “26%的性能提升对我们的HPC应用来说是很巨大的。Intel® Optane SSD将 大大帮助我们加速研究进程。”，赵勋旺, 陕西省超大规模电磁计算重点实验 室副主任 ATAGLANCE Hardware: Intel® Xeon® Gold 6148 processor Intel® Optane™ Solid State Drive DC P4800X Platform Features: Intel® Mesh Architecture More cores More memory channels XDU MOM-LASPCEM* increased performance with with Intel® Xeon® Gold 6148 processors and Intel® Optane™ SSD. 1 - Testing conducted on MOM-LASPCEM* software comparing 2S Intel® Xeon® Gold 6148 processor and Intel® Optane™ SSD to 2S Intel® Xeon® Gold 6148 processor and Intel® SSD DC P3700 Testing by Intel. For complete testing configuration details, see the Configuration Details section (slide 75). NormalizedPerformance www.xidian.edu.cn Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable ProcessorsHPC atomistic simulation Intel® optane™ Up to 1.26X faster with Intel® Optane™ SSD
- 99.
- 100. GENERATION- TO-GENERATION 100 Intel®Xeon®ProcessorScalableFamily Software Performance ProofPoints – HPC (Open Source) Potential Customer Benefits: • Faster time to completion for improved TCO • Higher resolution simulations for greater accuracy • More sophisticated analysis for higher confidence in outcomes hpc INTEL® AVX-512 MULTI- GENERATION 2.38XUp to Up to 1.85X Up to3.1X 1 1 – As demonstrated by proof points in this presentation Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others 1 1 Average Gen-to-Gen processor application speedup: 1.57X Average Multi Gen processor application speedup: 2.05X See slide 107 See slide 107 See slide 107
- 101.
- 102. 0 1 2 2S Intel® Xeon®processor E5-2697 v3 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 39% faster Workload: LAMMPS CG Water Simulation. 102 LAMMPS* life sciences Application: LAMMPS is a classical molecular dynamics code, and an acronym for Large-scale Atomic/Molecular Massively Parallel Simulator. It is used to simulate the movement of atoms to develop better therapeutics, improve alternative energy devices, develop new materials, and more. Potential Customer Benefits: The improved performance allows for longer time scales, larger simulations, and/or improved sampling and statistics. The continued advances in molecular dynamics performance on Intel® architecture allow computational scientists to solve new and more complex problems. Performance Factors: Intel® AVX-512 – Up to 49% gain1 versus Intel® AVX2. “The advances by Intel® corporation really enabled us to do simulations and address problems we could not before.” Valeria Molinero, Professor, University of Utah ATAGLANCE Hardware: Intel® Xeon® Gold 6148 Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® MPI Intel® Parallel Studio XE 2017 Composer Edition (C++/Fortran) Intel® Math Kernel Library multi-gen Up to 2.4X faster LAMMPS* increased performance1 with the 2S Intel® Xeon® Gold 6148 NormalizedPerformance lammps.sandia.gov Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Intel® AVX-512 1 - Testing conducted on LAMMPS* code comparing 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v3 and to 2S Intel® Xeon® Processor E5-2697 v4. Reported Intel® AVX-512 gains are compared to running an Intel® AVX2 binary using all cores on the same platform. Reported increased number of cores gains are compared to running reduced number of cores on the same platform. Testing done by Intel. - For complete testing configuration details, see the Configuration Details section, slide 135. 1.72X HPC
- 103. 103 GROMACS* Application: GROMACS is aversatile package to perform classical Molecular Dynamics simulations. Heavily optimized for most modern platforms and provides extremely high performance. Potential Customer Benefits: GROMACS provides users with wide range of functionality for chemical simulations. The improved performance on Intel® Xeon® new generation processors allows faster simulation of large biochemical systems due to enabling new AVX512 features and enabling enhanced parallelism. Performance Factors: Intel® AVX-512 ISA enabled by AVX512-intrinsics Intel® AVX-512 – Up to 21% gain1 versus Intel® AVX2 optimized baseline Multi-core - Up to 20% gain1 versus previous generations due to increased number of cores ATAGLANCE Hardware: Intel® Xeon® Gold 6148 processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® MPI Intel® Parallel Studio XE 2017 Composer Edition (C++/Fortran) Intel® Math Kernel Library Up to 2.4X faster www.gromacs.org Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Intel® AVX-512 1.29 1.27 1.30 2.01 1.82 1.89 0 1 2 water_1.5M_pme water_1.5M_rf lignocellulose_3M_rf 2S Intel® Xeon® processor E5-2697 v3 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor NormalizedPerformance Testcase: GROMACS workloads Increased GROMACS* performance1 with the 2S Intel® Xeon® Gold 6148 Up to 2X faster multi-gen 1 - Testing conducted on GROMACS* code comparing 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v3 and to 2S Intel® Xeon® Processor E5-2697 v4. Reported Intel® AVX-512 gains are compared to running an Intel® AVX2 binary using all cores on the same platform. Reported increased number of cores gains are compared to running reduced number of cores on the same platform. Testing done by Intel. - For complete testing configuration details, see the Configuration Details section, slide 135. 1.55X life sciencesHPC
- 104. 0 1 GaAsBi-64 Si256 2S Intel®Xeon® processor E5-2699 v3 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 Processor 104 VASP* Application: The Vienna Ab initio Simulation Package (VASP) is a computer program for atomic scale materials modeling and performs electronic structure calculations and quantum-mechanical molecular dynamics from first principles. VASP provides scientists with fast and precise calculation of materials properties covering wide range of MD methods from DFT, DFT- HF to Random-Phase approximation (GW, ACDFT). Potential Customer Benefits: VASP is one of the common compute package used at many supercomputing centers Continued performance advances delivers a better user experience with solution of new and more complex problems. Performance Factors: Intel® AVX512 – up to 25% gain1 over AVX2 Up to 10% gain from more cores. ATAGLANCE Hardware: Intel® Xeon® 2S Gold 6148 processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® MPI Intel® Parallel Studio XE 2017 Composer Edition (C++ / Fortran) Intel® Math Kernel Library Up to YY% faster https://www.vasp.at/ Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Intel® AVX-512 VASP Workloads:, GaAsBi-64(64 ions), Si256 (255 ions) multi-gen Increased GROMACS* performance1 with the 2S Intel® Xeon® Gold 6148 NormalizedPerformance 1.58X faster 1.9X faster 1 - Testing conducted on VASP* code comparing 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2699 v3 and to 2S Intel® Xeon® Processor E5-2697 v4. Reported Intel® AVX-512 gains are compared to running an Intel® AVX2 binary using all cores on the same platform. Reported increased number of cores gains are compared to running reduced number of cores on the same platform. Testing done by Intel. - For complete testing configuration details, see the Configuration Details section, slide 135. 1.58X 1.2X faster life sciencesHPC
- 105. 0 1 Cellulose NVE STMV 2SIntel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 Processor 105 Amberpme* Application: Amber* is a suite of programs for classical molecular dynamics and statistical analysis. The main MD program is PMEMD (Particle Mesh Ewald Molecular Dynamics) employs two separate algorithms for implicit- and explicit-solvent dynamics. Here performance for explicit solvent (PME) is presented. Larger PME benchmarks are benefitted more from 2S Intel® Xeon® Gold 6148. Potential Customer Benefits: Amber primarily used to investigate structures, fluctuations and thermodynamics of bio-molecular systems such as nucleic acids, proteins and cellular membranes Increased performance allows scientists to simulate more complex processes with accelerated time to solution and enhance drug discovery efforts. Performance Factors: Up to 73% performance increase1 from more cores, better memory bandwidth, and Intel® AVX-512 compared to the previous processor generation. ATAGLANCE Hardware: Intel® Xeon® Gold 6148 Processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) more cores, increased memory bandwidth Software Tools/Libraries: Intel® MPI Intel® Parallel Studio XE 2017 Composer Edition (C++/Fortran) More info http://ambermd.org/intel/ www.ambermd.org Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Intel® AVX-512 Amber* Explicit Solvent (PME) increased performance1 with the 2S Intel® Xeon® Gold 6148 processor Workload: Amber PME Cellulose NVE and STMV Simulation. NormalizedPerformance Up to 1.49X faster Up to 1.73X faster 1 - Testing conducted on Amber* code comparing 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v4. Reported Intel® AVX-512 gains are compared to running an Intel® AVX2 binary using all cores on the same platform. Reported increased number of cores gains are compared to running reduced number of cores on the same platform. Testing done by Intel. - For complete testing configuration details, see the Configuration Details section, slide 135. life sciencesHPC
- 106. 0 1 Rubisco Nucleosome 2S Intel®Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 Processor 106 Ambergb* Application: Amber* is a suite of programs for classical molecular dynamics and statistical analysis. The main MD program is PMEMD (Particle Mesh Ewald Molecular Dynamics) employs two separate algorithms for implicit (Generalized Born or GB)- and explicit-solvent (PME) dynamics. PMEMD is written in Fortran 90 and mainly MPI* parallelized; recently OpenMP*, mixed precision, and Vectorization added. This slide presents the implicit-solvent (GB) performance. Potential Customer Benefits: Amber primarily used to investigate structures, fluctuations and thermodynamics of bio-molecular systems such as nucleic acids, proteins and cellular membranes Increased performance allows scientists to simulate more complex processes with accelerated time to solution and enhance drug discovery efforts. Performance Factors: Up to 75% performance increase1 from more cores, better memory bandwidth, and Intel® AVX-512 compared to the previous processor generation. ATAGLANCE Hardware: Intel® Xeon® Gold 6148 Processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) more cores, increased memory bandwidth Software Tools/Libraries: Intel® MPI, Intel® Parallel Studio XE 2017 Composer Edition (C++/Fortran) More info http://ambermd.org/intel/ http://www.ambermd.org/ Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Intel® AVX-512 Amber* Implicit Solvent (GB) increased performance1 with the 2S Intel® Xeon® Gold 6148 processor Workload: Amber GB Nucleosome and Rubisco Simulation. Up to 1.67X faster Up to 1.75X faster 1 - Testing conducted on Amber* code comparing 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697v4. Reported Intel® AVX-512 gains are compared to running an Intel® AVX2 binary using all cores on the same platform. Reported increased number of cores gains are compared to running reduced number of cores on the same platform. Testing done by Intel. - For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance life sciencesHPC
- 107. 0 1 apoa1 stmv 2S Intel®Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 Processor 107 NAMD* Application: NAMD, recipient of a 2002 Gordon Bell Award, is a parallel molecular dynamics code designed for high-performance simulation of large biomolecular systems. Based on Charm++ parallel objects, NAMD scales to hundreds of cores for typical simulations and beyond 200,000 cores for the largest simulations. Potential Customer Benefits: NAMD is an important tool for molecular biophysics, among other things, to develop a model for the interaction of molecules in the body that allows for the development of new therapeutics for disease. Increased performance allows scientists to simulate more complex processes to enhance drug discovery efforts. Performance Factors: Intel® AVX-512 – Up to 42% gain versus Intel® AVX2. Up to 9.5% gain from more cores. ATAGLANCE Hardware: Intel® Xeon® Gold 6148 processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® MPI Intel® Parallel Studio XE 2017 Composer Edition (C++/Fortran) http://www.ks.uiuc.edu/Research/namd/ Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Intel® AVX-512 Testcase: apoa1(92K atoms), stmv(1M atoms) 1.67X faster1.54X faster NAMD* increased performance1 with the 2S Intel® Xeon® Gold 6148 NormalizedPerformance 1 - Testing conducted on NAMD* software comparing 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v4. Reported Intel® AVX-512 gains are compared to running an Intel® AVX2 binary using all cores on the same platform. Reported increased number of cores gains are compared to running reduced number of cores on the same platform. Testing done by Intel. - For complete testing configuration details, see the Configuration Details section, slide 135. life sciencesHPC
- 108. 0 1 NiO-64 C-64 Be-64NiO-32 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor NormalizedPerformance Increased QMCPACK* performance1 with the 2S Intel® Xeon® Gold 6148 Workloads: QMC samples/sec. NiO-64 (768 elec.), C-64 (256 elec.), Be-64 (256 elec.), NiO-32 (384elec) 108 QMCPACK* Application: QMCPACK is a US-DOE flagship many-body ab initio Quantum Monte Carlo code for computing the electronic structure of atoms, molecules, and solids. Workloads are selected to evaluate the performance vs problem sizes, physics and specialized methods in collaboration with QMCPACK Exascale Computing Project team. Potential Customer Benefits: Collaboration of ANL, Sandia and Intel® through Intel® Parallel Computing Center (IPCC). Continued performance advances increase the productivity of the developers and users Performance Factors: Portable code transformation to increase SIMD efficiency and reduce memory footprint. The efforts result in 3.1X-4.3X improvements on 2S Intel® Xeon® processor Gold 6148 and 2.6X-6.3X improvements on 2S Intel® Xeon® processor E5- 2697 v4 from baseline code (these results not in chart). ATAGLANCE Hardware: Intel® Xeon® 2S Gold 6148 processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® MPI Intel® Parallel Studio XE 2017 Composer Edition (C++/Fortran) Intel® Math Kernel Library Up to YY% faster http://qmcpack.org/ Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Intel® AVX-512 More info • Embracing a new era of highly efficient and productive quantum Monte Carlo simulations – submitted to SC17 • Optimization and parallelization of B-spline based orbital evaluations in QMC on multi/many-core shared memory processors Up to 45% faster 1 - Testing conducted on QMCPACK* code comparing 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v4. Reported Intel® AVX-512 gains are compared to running an Intel® AVX2 binary using all cores on the same platform. Reported increased number of cores gains are compared to running reduced number of cores on the same platform. Testing done by Intel. - For complete testing configuration details, see the Configuration Details section, slide 135. life sciencesHPC
- 109.
- 110. 0 1 2 3 2S Intel® Xeon®processor E5-2697 v3 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 1.3X faster Performance Metric: Speed-up using options/sec 110 MontECarloEuropeanoptions Application: Monte Carlo is a numerical method that uses statistical sampling techniques to approximate solutions to quantitative problems. In finance, Monte Carlo algorithms are used to evaluate complex instruments, portfolios, and investments. This is compute bound, double precision workload. Potential Customer Benefits: Higher performance allow either doing the same work faster leading to improved TCO or simulation of more paths leading to higher confidence in results. Performance Factors: Using Intel® AVX-512 SIMD vectorization improved performance by 1.85X over Intel® AVX2. Higher core counts of Intel Xeon® Gold 6148 processor contributes to higher performance. Better memory hierarchy adds to the performance Code modernization strategy: Parallelizing outer loop over options and vectorize inner loop of paths. ATAGLANCE Hardware: 2S Intel® Xeon® Gold 6148 Processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) More cores Improved memory hierarchy Software Tools/Libraries: Intel® Parallel Studio XE 2017 Composer Edition (C++) Monte Carlo European Option increased performance with the 2S Intel® Xeon® Gold 6148 processor1 1 - Testing conducted on Monte Carlo software comparing 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v3 and to 2S Intel® Xeon® Processor E5-2697 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Up to 3.1X faster Intel® AVX-512 multi-gen 2.38X FinancialservicesHPC
- 111. 0 1 2 2S Intel® Xeon®processor E5-2697 v3 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 1.34X faster Performance Metric: Speed-up using options/sec 111 BlackSCHOLES Application: Black-Scholes is a popular mathematical model used in finance for European option valuation. This is a double precision version. Potential Customer Benefits: The improvement in performance provides improved TCO. Better performances and optimizations enables generation of more accurate financial models. Performance Factors: Intel® AVX-512 contributed a 1.35X gain compared to Intel® AVX2. Combined parallelization and vectorization. Higher memory bandwidth adds to the performance. ATAGLANCE Hardware: 2S Intel® Xeon® Gold 6148 processor Platform Features: Multi core, Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Improved memory hierarchy Software Tools/Libraries: Intel® Parallel Studio XE 2017 Composer Edition (C++) Black-Scholes increased performance with the 2S Intel® Xeon® Gold 6148 processor1 1 - Testing conducted on Black-Scholes software comparing 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v3 and to 2S Intel® Xeon® Processor E5-2697 v4. Testing done by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Up to 2.5X faster Intel® AVX-512 multi-gen 1.86X FinancialservicesHPC
- 112. 0 1 2S Intel® Xeon®processor E5-2697 v3 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 1.37X faster 112 Binomialoptionpricing Application: Binomial option pricing is a lattice-based approach that uses a discrete-time model of the varying price over time of the underlying financial instrument. This is compute bound, double precision workload. Potential Customer Benefits: The improved performance allows for larger simulations and more accurate results. Performance Factors: Intel® AVX-512 SIMD vectorization improved performance by 1.33X compared to Intel® AVX2. Higher core counts of Intel Xeon® Gold 6148 processor contributes to higher performance Code modernization strategy: Parallelizing outer loop over options and vectorizing inner loop over the grid. ATAGLANCE Hardware: 2S Intel® Xeon® Gold 6148 processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) More cores Improved memory hierarchy Software Tools/Libraries: Intel® Parallel Studio XE 2017 Composer Edition (C++) Increased Binomial Option Pricing performance with the 2S Intel® Xeon® Gold 6148 processor1 NormalizedPerformance Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Up to 1.95X faster Intel® AVX-512 multi-gen 1 - Testing conducted on Binomial Option Pricing code comparing 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v3 and to 2S Intel® Xeon® Processor E5-2697 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. Workload: Binomial double precision workload 1.42X FinancialservicesHPC
- 113. physics Unlock, discover, innovate.Simulations – Fast, detailed, accurate. geophysics
- 114. 0 1 2 2S Intel® Xeon®processor E5-2697 v4 2S Intel® Xeon® 6148 Gold processor Workload: BerkeleyGW 1.2 MTXEL-COR phase Sigma Benzene Simulation 114 NERSCBerKELEYGW Application: BerkeleyGW is a massively parallel computational package for electron excited state properties that is based on many-body perturbation theory employing the ab initio GW and GW plus Bethe-Salpeter equation methodology. Sigma is the second half of the GW code. It gives the quasiparticle self-energies and dispersion relation for quasielectron and quasihole states. Potential Customer Benefits: More efficient single node performance reduces overall communication cost of massive scaling same-sized problem. Improved performance enables analysis of larger workloads. expanding the range of materials which can be studied. Performance Factors: Using Intel® AVX-512 SIMD vectorization reduced runtime 29% relative to Intel® AVX2. ATAGLANCE Hardware: Intel® Xeon® Gold 6148 processor Platform Features:. Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel Parallel Studio XE Intel Math Kernel Library Intel® MPI Up to 2.12X faster BerkeleyGW increased performance with the Intel® Xeon® Gold 6148 processor1 1 - Testing conducted on BerkeleyGW* software comparing 2S Intel® Xeon® 6148 Gold processor to 2S Intel® Xeon® Processor E5-2697 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance Berkeleygw.org Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable ProcessorsSOLIDSTATEPHYSICSHPC
- 115. 0 1 2S Intel® Xeon®processor E5-2697 v3 2S Intel® Xeon® processor E5-2699 v4 2S Intel® Xeon® Gold 6148 processor Up to 1.32X faster Up to 1.95X faster 115 ATAGLANCE Hardware: Intel® Xeon® Gold 6148 processor Platform Features: More cores More memory bandwidth Software Tools/Libraries: Intel® MPI Intel® Parallel Studio XE 2017 Composer Edition (C++/Fortran) Up to YY% faster www.pppl.gov Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable ProcessorsPlasmaPhysics multi-gen Application: GTC-P is a proxy code for plasma physics workloads used to model Tokamak reactors in development. It is a particle-in-cell code, using a Lagrangian representation of the helical path ions take through the reactor and a Eulerian description of the background plasma. Potential Customer Benefits: Optimizations in this code will be used to drive optimizations in the scientific code, and serve as a template for other PIC and plasma codes. Performance Factors: More cores, more memory bandwidth, and improved microarchitecture of the Intel® Xeon® Scalable processor platform contributed to the improved performance. PPPLGTC-P* 1 - Testing conducted on GTC-P comparing 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v3 and 2S Intel® Xeon® Processor E5-2699 v4. Testing done by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. Workload: B-1rank-half.txt, with 400 particles/cell NormalizedPerformance PPPL GTC-P* increased performance with the 2S Intel® Xeon® Gold 61481 1.47X HPC
- 116. 0 1 2S Intel® Xeon®processor E5-2699 v3 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 25% faster 116 MILC Application: MILC software represents a set of codes written by the MIMD Lattice Computation (MILC) collaboration used to study quantum chromodynamics (QCD), the theory of the strong interactions of subatomic physics. It performs simulations of four-dimensional SU lattice gauge theory on MIMD (Multiple Instruction, Multiple Data) parallel machines. Potential Customer Benefits: MILC applications address fundamental questions in high energy and nuclear physics, and is directly related to major experimental programs in these fields MILC is one of the largest compute cycle users at many U.S. and European supercomputing centers Performance Factors: Enabled QPhiX* Library: Provides sparse solvers and Dslash kernels for Lattice QCD simulations optimized for Intel® architectures. Memory bandwidth improvements speedup application performance by 1.9X1. The application is memory bandwidth bound. ATAGLANCE Hardware: Intel® Xeon® 6148 Gold processor Platform Features: Memory Bandwidth Software Tools/Libraries Intel® Parallel Studio XE 2017 Cluster Edition update 2 Intel® MPI QPhiX Library MILC increased performance with the Intel® Xeon® 6148 Gold processor More info Performance Recipe - here NormalizedPerformance Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors http://physics.indiana.edu/~sg/milc.html Quantum chromodynamics Workload: ks_imp_rhmc, Lattice Size: 24 ^ (4) multi-gen Up to 90% faster 1 - Testing conducted on MILC* software comparing 2S Intel® Xeon® 6148 Gold processor to 2S Intel® Xeon® processor E5-2699 v3 and to 2S Intel® Xeon® processor E5-2697 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. 1.52X HPC
- 117. 0 1 2S Intel® Xeon®processor E5-2697 v3 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 24% faster 117 SPECMPI2007 Benchmarksuite Application: SPEC MPI® 2007 benchmark suite is for evaluating MPI-parallel, floating point, compute intensive performance across a wide range of cluster and SMP hardware. Potential Customer Benefits: SPEC gives users the most objective and representative benchmark suite for measuring and comparing high- performance computer systems. Performance Factors: Intel® AVX-512 contributes up to 18% performance boost per component. Memory bandwidth improvements speedup application performance up to 71% in geomean. Memory bandwidth contributes to performance increase. ATAGLANCE Hardware: Intel® Xeon® Gold 6148 processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® MPI, Intel® Parallel Studio XE Composer (C++/Fortran) multi-gen Up to 71% faster Increased SPEC MPI 2007 performance with the 2S Intel® Xeon® Gold 61481 NormalizedPerformance www.spec.org/mpi2007 Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Intel® AVX-512 1 - Testing conducted on SPEC MPI software comparing 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® processor E5-2697 v3 and to 2S Intel® Xeon® processor E5-2697 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. Workload: geomean over 13 apps 1.37X HPC
- 118. 0 1 2S Intel® Xeon®processor E5-2697 v4 2S Intel® Xeon® 6148 Gold processor 118 HOMME Application: HOMME is the spectral element dynamical core that solves the equations of motion in the CAM-SE atmospheric model, part of the NSF Community Earth System Model (CESM) as well as the related DOE ACME model. CESM is a widely-used Earth system model and an important source of simulations used by the Intergovernmental Panel on Climate Change. Potential Customer Benefits: Faster execution allows researchers to run longer simulations and more ensemble members, enabling study of more climate change scenarios and greater confidence in predictions. Ability to push many atmospheric tracers in reasonable time allows detailed representation of high-altitude chemical processes such as stratospheric ozone loss. Performance Factors: Improved memory bandwidth speeds up bandwidth-limited Eulerian tracer advection. ATAGLANCE Hardware: Intel® Xeon® 6148 Gold Processor Platform Features: Increased memory bandwidth Software Tools/Libraries: Intel Parallel Studio XE 2017 Cluster Edition Update 2 Up to 1.67X faster HOMME increased performance with the Intel® Xeon® Gold 6148 processor1 1 - Testing conducted on HOMME* software comparing 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® processor E5-2697 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable ProcessorsEARTH SYSTEMMODELS https://www.homme.ucar.edu/ Workload: Baroclinic instability simulation in a “whole atmosphere” (extending to lower thermosphere) configuration HPC
- 119. 0 1 Base Peak 2S Intel®Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 Processor 119 OMP2012 Application: The successor to the OMP2001, designed for measuring performance using applications based on the OpenMP 3.1 standard for shared-memory parallel processing. OMP2012 also includes an optional metric for measuring energy consumption. Potential Customer Benefits: SPEC gives users the most objective and representative benchmark suite for measuring and comparing high- performance computer systems. Performance Factors: Improved memory bandwidth together with wider vector unit and more cores speed up suite applications. ATAGLANCE Hardware: Intel® Xeon® 6148 Gold processor Platform Features:. Increased memory bandwidth Wider vector unit Increased number of cores Intel SW Tools/Libraries: Intel® Parallel Studio XE Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable ProcessorsStandard PerformanceEvaluation https://www.spec.org/ Up to 1.46X faster SPEC OMP2012 increased performance with the Intel® Xeon® Gold 6148 processor1 NormalizedPerformance Workload: geomean over 14 apps 1 - Testing conducted on SPEC OMP2012* software comparing 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® processor E5-2697 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. HPC
- 120. 0 1 2S Intel® Xeon®processor 2699 v3 2S Intel® Xeon® processor 2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 11% faster Workload: H2O-64 120 CP2K* OpenSource MolecularDynamics Application: CP2K is a powerful and scalable program for atomistic simulations of a wide range of systems. CP2K features a wide range of atomistic interaction models incl. classical potentials, semi-empirical schemes, Density Functional Theory, Hartree- Fock (HF), and post-HF correlation methods such as MP2 and RPA. Potential Customer Benefits: Speed up for (highly) parallelized as well as (partially) sequential workloads or memory bandwidth-bound workloads. Higher throughput or faster time to solution enables more detailed or more complex analysis types. Platform Factors: LIBXSMM: Intel’s highly optimized Open Source Library for small, dense or sparse matrix multiplication, and small convolutions. Intel® AVX-512. ATAGLANCE Hardware: Intel® Xeon® Gold 6148 processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Higher multicore performance Higher per-core performance Software Tools/Libraries): Intel® Math Kernel Library Intel® Compiler LIBXSMM Intel® AVX-512 Up to 36% faster CP2K* increased performance with the Intel® Xeon® Gold 6148 processor1 1 - Testing conducted on CP2K* software comparing the 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2699 v3 and to 2S Intel® Xeon® Processor E5-2697 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance Open Source Molecular Dynamics www.cp2k.org Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others multi-genIntel® Xeon® Scalable Processors 1.22X HPC atomistic simulation
- 121. 0 1 2S Intel® Xeon®processor 2699 v3 2S Intel® Xeon® processor 2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 6% faster Workload: LIBTEST 121 Application: CP2K is a powerful and scalable program for atomistic simulations of a wide range of systems. CP2K features a wide range of atomistic interaction models incl. classical potentials, semi-empirical schemes, Density Functional Theory, Hartree- Fock (HF), and post-HF correlation methods such as MP2 and RPA. Potential Customer Benefits: Speed up for (highly) parallelized as well as (partially) sequential workloads, or memory bandwidth-bound workloads. Higher throughput or faster time to solution enables more detailed or more complex analysis types. Performance Factors: LIBXSMM: Intel’s highly optimized Open Source Library for small, dense or sparse matrix multiplication, and small convolutions. Intel® AVX-512 provided an up to 55% gain compared to Intel AVX2 (both on Intel® Xeon® Gold 6148 processor) Intel® AVX-512 Up to 87% faster CP2K* increased performance with the Intel® Xeon® Gold 6148 processor1 1 - Testing conducted on CP2K* software comparing the Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2699 v3 and to 2S Intel® Xeon® Processor E5-2697 v4. Testing done by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others multi-gen Open Source Molecular Dynamics www.cp2k.org ATAGLANCE Hardware: Intel® Xeon® Gold 6148 processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Higher multicore performance Higher per-core performance Software Tools/Libraries): Intel® Math Kernel Library Intel® Compiler LIBXSMM Intel® Xeon® Scalable Processors CP2K* OpenSource MolecularDynamics 1.76X HPC atomistic simulation
- 122. 0 1 2 2S Intel® Xeon®processor 2699 v3 2S Intel® Xeon® processor 2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 1.57X faster Workload: DFT-LS (NREP=3) 122 CP2K is a powerful and scalable program for atomistic simulations of a wide range of systems. CP2K features a wide range of atomistic interaction models incl. classical potentials, semi-empirical schemes, Density Functional Theory, Hartree- Fock (HF), and post-HF correlation methods such as MP2 and RPA. Potential Customer Benefits: Speedup for (highly) parallelized as well as (partially) sequential workloads, or memory bandwidth-bound workloads. Higher throughput or faster time to solution enables more detailed or more complex analysis types. Performance Factors: LIBXSMM: Intel’s highly optimized Open Source Library for small, dense or sparse matrix multiplication, and small convolutions. Intel® AVX-512 provided an up to 30% gain compared to Intel AVX2 (both on Intel® Xeon® Gold 6148 processor) Intel® AVX-512 Up to 2.43X faster CP2K* increased performance with the Intel® Xeon® Gold 6148 processor1 1 - Testing conducted on CP2K* software comparing the Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2699 v3 and to 2S Intel® Xeon® Processor E5-2697 v4. Testing done by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others multi-gen Open Source Molecular Dynamics www.cp2k.org ATAGLANCE Hardware: Intel® Xeon® Gold 6148 processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Higher multicore performance Higher per-core performance Software Tools/Libraries): Intel® Math Kernel Library Intel® Compiler LIBXSMM Intel® Xeon® Scalable Processors CP2K* OpenSource MolecularDynamics 1.54X HPC atomistic simulation
- 123. 0 1 2S Intel® Xeon®processor E5-2697 v3 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 1.36X faster 123 ATAGLANCE Hardware: Intel® Xeon® 6148 Gold processor Platform Features: More flops Larger core count Intel® Hyper-Threading Technology Higher memory bandwidth Software Tools/Libraries: Intel® C++ Compiler Up to 1.71X faster SPECFEM3D_GLOBE Performance Improvement with the 2S Intel® Xeon® 6148 Gold processor1 NormalizedPerformance Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors multi-gen Specfem3d_globe* Application: SPECFEM3D_GLOBE simulates the three-dimensional global and regional seismic wave propagation based upon the spectral- element method (SEM). It is a time-step algorithm which simulates the propagation of earth waves given the initial conditions, mesh coordinates/ details of the earth crust. Potential Customer Benefits: Intel® Xeon® processor continues to improve performance for scalable workloads. SPECFEM3D_GLOBE benefits from more cores and higher- bandwidth memory available on the latest Intel® Xeon® processor. Performance Factors: Larger core count/ threading. Higher DDR memory bandwidth. Out-of-the-box performance gains. Workload: 14000 steps, small_benchmark_run_to_test_more_complex_Earth Computational Seismology 1 - Testing conducted on SPECFEM2D_GLOBE* software comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v3 and to 2S Intel® Xeon® Processor E5-2697 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. 1.25X HPC
- 124. 0 1 2S Intel® Xeon®processor E5-2697 v3 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 1.09X faster 124 PhysicsQCD Application: QPhiX is an optimized solver library for QCD on Intel® Xeon® and Xeon Phi™ processors and provides implementation for Dslash operator and CG, BICGStab and mixed precision solvers for Wilson and Clover improved Wilson Quarks. Potential Customer Benefits: Lattice calculations are an important component of the nuclear physics research. QPhiX helps speed up the computation by multiple folds on Intel processors. Intel® Xeon® processor further improves performance with features such as higher memory bandwidth vs. previous generation, AVX-512 vector instruction set architecture. Performance Factors: Larger core count/ threading. Transparent use of higher DDR memory bandwidth. Intel® AVX-512 contributed to improved performance. Up to 1.61X faster QPhiX CG solver performance improved with Intel® Xeon® Gold 6148 processor1 NormalizedPerformance Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others multi-gen QPHIX Workload: Single precision QPhiX Benchmark (time_dslash_noqdp), QUDA (dslash_test) 32*32*32*64 ATAGLANCE Hardware: Intel® Xeon® 6148 Gold processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Intel® Hyper-Threading Technology More flops Larger core count Higher memory bandwidth Software Tools/Libraries: Intel® C++ Compiler Intel® AVX-512 1 - Testing conducted on QphiX* software comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v3 and to 2S Intel® Xeon® Processor E5-2697 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. Intel® Xeon® Scalable Processors 1.47X HPC
- 125. 0 1 2S Intel® Xeon®processor E5-2697 v3 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor Up to 1.11X faster Workload: Chroma hmc double precision 24*24*24*64 125 Application: The Chroma package supports data-parallel programming constructs for lattice field theory and in particular lattice QCD. It uses the SciDAC QDP++ data-parallel programming (in C++) that presents a single high-level code image to the user, but can generate highly optimized code for many architectural systems including single node workstations, multi and many- core nodes, clusters of nodes via QMP, and classic vector computers. Potential Customer Benefits: Chroma is deployed on numerous supercomputers and one of the most used QCD applications/ research kernels. Intel’s optimizations are incorporated into mainline Chroma. The optimizations are made available in the QPhiX library. Performance Factors: Larger core count/ threading. Higher DDR memory bandwidth. Intel® AVX-512 helped improve performance. ATAGLANCE Hardware: Intel® Xeon® 6148 Gold processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Intel® Hyper-Threading Technology More flops Larger core count Higher memory bandwidth Software Tools/Libraries: Intel® C++ Compiler Up to 1.6X faster Chroma* BiCGStab performance improved with Intel® Xeon® Gold 6148 processor1 NormalizedPerformance Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others multi-gen Chroma* 1 - Testing conducted on Chroma* software comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® Processor E5-2697 v3 and to 2S Intel® Xeon® Processor E5-2697 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. Intel® AVX-512 Intel® Xeon® Scalable Processors 1.44X PhysicsQCDHPC
- 126. 0 1 2S Intel® Xeon®processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor 126 Argonnenationallab(hacc)* Brief Application / Workload Description: HACC (Hardware Accelerated Cosmology Code) is particle- mesh code designed to simulate astrophysical effects to find the cause of the accelerated expansion of the universe. A 2563 grid is used as input workload Potential Customer Benefits: Simulate larger problem sizes Simulate physics at a finer granularity to make results more accurate Performance Factors: This application is enabled with both MPI and OpenMP parallelization. Higher core counts of Intel Xeon® Gold 6148 processor leads to higher performance. ATAGLANCE Hardware: Intel® Xeon® 6148 Gold processor Platform Features: More cores Improved memory bandwidth Software Tools/Libraries: Intel® ICC Compiler Intel® Math Kernal Library Up to 15% faster HACC* increased performance with the Intel® Xeon® Gold 6148 processor1 1 - Testing conducted on ANL HACC* software comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® processor E5-2697 v4 Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. NormalizedPerformance www.alcf.anl.gov Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable ProcessorsCOSMOLOGY Workload: 2563 N-body particles HPC
- 127. EnergyIndustry Enhancing exploration andextraction processes.
- 128. 0 1 2 2S Intel® Xeon®processor E5-2680 v3 2S Intel® Xeon® processor E5-2680 v4 2S Intel® Xeon® Gold 6148 processor 128 MinesparistechFwi2d-inversion Application: FWI2D-inversion is a 2D time-domain Full Waveform inversion that uses finite differences for both forward and backward problems and a quasi-Newton (with L-BFGS algorithm) optimization scheme for the model parameters update. It is based on a 2D isotropic acoustic wave equation with variable density and P-wave velocity and uses a 8th order in space (symmetric coefficient) and 4th order in time scheme with absorbing boundary conditions. Potential Customer Benefits: The improved performance reduces time to solution of current simulations and allow frequency increase for more accurate imaging. The continued advances in seismic modeling performance on Intel® architecture allow computational scientists improvements to produce higher resolution images of the subsurface, or to improved throughput of their existing workload. Performance Factors: NO intrinsic. Parallelization is achieved through standard MPI shot gathers distribution and OpenMP for domain decomposition. Intel® AVX-512 and memory bandwidth help provide performance increase Up to 1.52X at node level. multi-gen Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Intel® AVX-512 ATAGLANCE Hardware: Intel® Xeon® Gold 6148 processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512), Multi-core Software Tools/Libraries: Intel® Parallel Studio XE 2017 Composer Edition (Fortran) ENERGY www.mines-paristech.eu FWI2D-Inversion* increased performance with the Intel® Xeon® Gold 6148 processor1 NormalizedPerformance Workload: Extended Marmousi-1 , single-node problem- size: 600x1423 Up to 2.41X faster Up to 1.3X faster 1 - Testing conducted on FWI2D software comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® processor E5-2680 v3 and to 2S Intel® Xeon® processor E5-2680 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. 1.85X HPC
- 129. 0 1 2 2S Intel® Xeon®processor E5-2680 v3 2S Intel® Xeon® processor E5-2680 v4 2S Intel® Xeon® Gold 6148 processor 129 MinesparistechFwi2d-modeling Application: FWI2D-modeling denotes the 2D time-domain forward modeling engine that uses finite-difference to generate synthetic seismic shot gathers for later use in the minimization process of the whole Full Waveform Inversion. It is based on a 2D isotropic acoustic wave equation with variable density and P-wave velocity and uses a 8th order in space (symmetric coefficient) and 4th order in time scheme with absorbing boundary conditions. Potential Customer Benefits: The improved performance reduces time to solution of current simulations and allow frequency increase for more accurate imaging. The continued advances in seismic modeling performance on Intel® architecture allow computational scientists improvements to produce higher resolution images of the subsurface, or to improved throughput of their existing workload. Performance Factors: NO intrinsic. Parallelization is achieved through standard MPI shot gathers distribution and OpenMP for domain decomposition. Intel® AVX-512 and memory bandwidth help provide performance increase Up to 1.45X at node level. multi-gen Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Intel® AVX-512 ATAGLANCE Hardware: Intel® Xeon® Gold 6148 processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® Parallel Studio XE 2017 Composer Edition (Fortran) www.mines-paristech.eu FWI2D-Modeling* increased performance with the Intel® Xeon® Gold 6148 processor1 NormalizedPerformance Workload: Extended Marmousi-1 , single-node problem- size: 600x1423 Up to 2.15X faster Up to 1.25X faster 1 - Testing conducted on FWI2D software comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® processor E5-2680 v3 and to 2S Intel® Xeon® processor E5-2680 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. 1.72X ENERGYHPC
- 130. 0 1 2S Intel® Xeon®processor E5-2680 v3 2S Intel® Xeon® processor E5-2680 v4 2S Intel® Xeon® Gold 6148 processor 130 iso3dfd Application: Iso3dfd is a finite difference kernel used to solve 3D acoustic isotropic wave equation for seismic modeling (Acquisition design) and imaging (RTM, FWI). This is a 16th order in space (symmetric coefficient) and 2nd order in time scheme without boundary conditions. Potential Customer Benefits: The improved performance reduces time to solution of current simulations and allow frequency increase for more accurate modeling & imaging. The continued advances in seismic modeling performance on Intel® architecture allow computational scientists to solve new and more complex problems. Performance Factors: NO intrinsic. Cache blocking optimized using Genetic algorithm. OMP 4.5 task. Intel® AVX-512 – Up to 58% gain versus Intel® AVX2. multi-gen Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors Intel® AVX-512 ATAGLANCE Hardware: Intel® Xeon® Gold 6148 processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® Parallel Studio XE 2017 Composer Edition (C/C++) ISO3DFD* increased performance with the Intel® Xeon® Gold 6148 processor1 NormalizedPerformance Workload: Homogeneous model , single-node problem- size: 224x212x2100 Up to 1.88X faster 1 - Testing conducted on ISO3DFD software comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® processor E5-2680 v3 and to 2S Intel® Xeon® processor E5-2680 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. 1.67X Up to 12% faster ENERGYHPC
- 131. 0 1 2S Intel® Xeon®processor 2699 v3 2S Intel® Xeon® processor 2697 v4 2S Intel® Xeon® Gold 6148 processor Workload: YASK (AWP), 1024*1024*128 grid points 131 YASKAWP* Brief Application / Workload Description: YASK: Yet Another Stencil Kernel; Intel open-source software framework for stencil-code development AWP: compute kernel used in Anelastic Wave Propagation— ODC earthquake simulator; see http://hpgeoc.sdsc.edu/AWPODC Potential Customer Benefits: Faster time-to-solution. Higher resolution simulation. Performance Factors: Intel® AVX-512 instruction set. Stencil “vector-folding” to maximize reuse. Intel® C++ Compiler. Multi-core with hyper-threading. Improved memory sub-system. ATAGLANCE Hardware: Intel® Xeon® Gold 6148 processor Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Software Tools/Libraries: Intel® C++ Compiler YASK open-source framework Up to 88% faster YASK (AWP)* increased performance with the Intel® Xeon® Gold 6148 processor1 1 - Testing conducted on YASK* software comparing Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® processor E5-2699 v3 and to 2S Intel® Xeon® processor E5-2697 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. More info https://software.intel.com/en- us/articles/recipe-building-and- running-yask-yet-another-stencil- kernel-on-intel-processors NormalizedPerformance 01.org/yask Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable Processors multi-genIntel® AVX-512 Up to 19% faster Y.A.S.K. 1.57X ENERGYHPC
- 132. ClimateAndWeather Increasing accuracy andtimeliness of forecasts.
- 133. 0 1 CONSUS 12km CONSUS2.5km 2SIntel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor NormalizedPerformance Increased WRF* performance1 with the 2S Intel® Xeon® Gold 6148 133 WRF* Application: The Weather Research and Forecasting (WRF) Model is a next- generation mesoscale numerical weather prediction system designed for both atmospheric research and operational forecasting needs. It features two dynamical cores, a data assimilation system, and a software architecture facilitating parallel computation and system extensibility. The model serves a wide range of meteorological applications across scales from tens of meters to thousands of kilometers. Potential Customer Impact: Faster execution allows researchers to run longer simulations and with more precision and with more prediction options. Performance Factors: Improved memory bandwidth speeds up bandwidth-limited tracer advection. ATAGLANCE Hardware: Intel® Xeon® 6148 Gold processor Platform Features:. Increased memory bandwidth Software Tools/Libraries: Intel® Parallel Studio XE Cluster Edition Up to 1.41X faster Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® Xeon® Scalable ProcessorsNumeric WeatherSimulation https://www.wrf-model.org/ Up to 1.35X faster Workload: CONUS12km and CONUS2.5km 1 - Testing conducted on WRF* software comparing 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® processor E5-2697 v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. HPC
- 134. 134 weather ATAGLANCE Hardware: Intel® Xeon® Gold6148 processor Intel® Solid State Drive Data Center S3500 Platform Features: Intel® Advanced Vector Extensions 512 (Intel® AVX-512) Intel® Omni-Path Architecture 1 - Testing conducted on MPAS-O software comparing 2S Intel® Xeon® Gold 6148 processor to 2S Intel® Xeon® E5-2697 processor v4. Testing by Intel. For complete testing configuration details, see the Configuration Details section, slide 135. https://mpas-dev.github.io/ Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. *Other names and brands may be claimed as the property of others Intel® AVX-512 MPASOCEAN4.0* Application: MPAS (Model for Prediction Across Scales) is a suite of programs for atmosphere, ocean, and other earth-system simulation. LANL is primarily responsible for the MPAS Ocean (MPAS-O) model. MPAS-O has demonstrated the ability to accurately reproduce mesoscale activity. The MPAS-O code is written in Fortran 90 and C++. The code is MPI* parallelized with full OpenMP* threading added. The OpenMP implementation needs testing. Potential Customer Benefits: Faster run times and more accurate data. Potential to include larger workloads for longer time scales. Performance Factors: By taking advantage of more cores, more memory, and Intel® AVX-512, developers generated an up to 37% performance improvement at 8 nodes compared to the previous processor generation.0 1 1 Node 4 Nodes 8 Nodes 2S Intel® Xeon® processor E5-2697 v4 2S Intel® Xeon® Gold 6148 processor NormalizedPerformance Increased GROMACS* performance1 with the 2S Intel® Xeon® Gold 6148 Workload: EC_60to30km run times for 2 day pure MPI simulation, excludes time to I/O Up to 37% faster Intel® Xeon® Scalable ProcessorsHPC
- 135.
- 136. * Other Namesand Brands May Be Claimed as the Property of Others 136 IncreasenOsqlreal-timeperformancewithaerospikeserver* andIntel®Xeon®Platinum8180Processor 2015 2017 Software Upgrade Aerospike Server Enterprise 3.12.1 Hybrid Memory Architecture database optimized for flash and SSD Software enhancements such as “auto tuning” delivered 2.1X transactions per second Automatically takes advantage of more cores & higher memory bandwidth of the latest Xeon® Intel® platform. Hardware Upgrade Intel® Xeon® Platinum 8180 Processor More cores – up to 28 – that benefit compute-heavy workloads More memory bandwidth with 6 DDR4 memory channels per socket Software Baseline Aerospike Server Enterprise 3.6.4 Better Together Up to 4X1 improvement in transactions per second More transactions processed in less time creates a better experience for Aerospike customers Hardware Baseline Intel® Xeon® processor E5-2697 v3 Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. 1 – For complete configuration details, see Configuration Details section, slide 135. Benchmark: The database was populated with 400 M records of 100 bytes each and benchmarked with the Aerospike Java Benchmark tool
- 137. * Other Namesand Brands May Be Claimed as the Property of Others 137 IncreaseanalyticqueriesandsupportmoreconcurrentuserswithIBM®DB2 andIntel®Xeon®Platinum8180Processor 2013 2017 Software Upgrade IBM® DB2 v11.1 1.22X gain1 from DB2 v10.5 using identical hardware using identical BDInsights workload In-memory BLU functionality and scalability improvements Bufferpool scalability improvements Hardware Upgrade Intel® Xeon® Platinum 8180 Processor 2.17x gain compared to Intel® Xeon® processor E7-4890 v2 using identical database software using identical BDInsights workload Increased parallelism - up to 28 cores available 1.5X memory bandwidth with 6 DDR4 memory channels2 Improved I/O bandwidth with up to 48 PCIe* 3.0 lanes available IBM® DB2 10.5 Better Together 2.65X1 improvement when upgrading software and hardware Lower response time for faster queries Higher throughput to support more concurrent usersIntel® Xeon® E7-4890 v2 Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. 1 – see Configuration Details section, slide 135. 2 – Total Memory capacity Benchmark: The proprietary IBM Big Data Insights Workload (BDInsights) is a multi-user data warehousing workload based on a retail environment. The workload as configured uses a 3TB scale factor, 12 concurrent users, and a mix of intermediate and complex analytics queries, and is compute-bound in the tested configuration. 2014
- 138. 138 learnmoreaboutIntel®Xeon®Scalableprocessors *Other names andbrands may be claimed as the property of others See more enabled software performance content at www.intel.com/XeonSoftwareSolutions Learn more about the Intel® Xeon® Scalable Processors at www.intel.com/XeonScalable
- 139.
- 140. 140 Virtuozzo Containers: SPECjbb+WebBenchworkload. OS: Virtuozzo Linux* 7 kernel 3.10.0-514.16.1.vz7.30.10 Testing by Virtuozzo, May 2017. BASELINE: 2S Intel® Xeon® processor E5-2680 v4, 2.4GHz, 14 cores, turbo and HT on, 256GB total memory, DDR4 DIMM, Intel® SSD. NEW: 2S Intel® Xeon® Platinum 8168 processor, 2.7GHz, 24 cores, turbo and HT on, 192 GB total memory, Intel® SSD. Huawei FusionSphere* virtualized cloud Platform: VMM OS:Fusionsphere 6.1, Guest OS: CentOS* 7.2. Testing by Huawei/Intel May 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, 256GB total memory, 16 slots / 16GB / 2400 MT/s / DDR4, Intel® SSD DC S3700 800GB. NEW: 2S Intel® Xeon® Platinum 8180 processor, 2.5GHz, 28 cores, turbo and HT on, 192GB total memory, 12 slots / 16GB / 2666 MT/s / DDR4, Intel® SSD DC S3700 800GB. Neusoft SaCa® Aclome*: SaCa® Aclome workload (for general performance) and compressing/decompressing workload (for QAT). OS: CentOS 7.3.1611. Testing by Intel and Neusoft May 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4, SATA SSD. NEW: 2S Intel® Xeon® Platinum 8180 processor, 2.5GHz, 28 cores, turbo and HT on, Intel® Chipset C627, 128GB total memory, 8 slots / 16GB / 2666 MT/s / DDR4, SATA SSD. Notes: Data compression/depression using Zlib 1.2.8. QAT Driver version: S4PR1-Linux-QAT1.7.Upstream.L.0.9.0-36 Zizo: OS: Ubuntu Linux* 16.04 kernel 4.4. Testing by Zizo May 2017. BASELINE: 4S Intel® Xeon® processor E5-2680 v4, 2.4GHz, 14 cores, turbo and HT on, 256GB total memory, DDR4 DIMM, Intel SSD. NEW: 4S Intel® Xeon® Platinum processor 8168 v4, 2.7GHz, 24 cores, turbo and HT on, 192 GB total memory, Intel SSD. Baidu Search Click-Through-Rate (CTR)* : OS: CentOS Linux release 7.3.1611. Testing by Intel June 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4 RDIMM, Intel® SSD DC S3700 800GB. NEW: Intel® Xeon® Platinum 8180 processor, 2.5GHz, 28 cores, turbo and HT on, 192GB total memory, 12 slots / 16GB / 2666 MT/s / DDR4 RDIMM, Intel® SSD DC S3700 800GB. Tencent Business Analytics*: Video Stitching workload. OS: CentOS 7.3.1611 Linux kernel 4.9.8. Testing by Intel April 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4, SATA3 4TB. NEW: Intel® Xeon® Platinum 8180 processor, 2.5GHz, 28 cores, turbo and HT on, 192GB total memory, 12 slots / 16GB / 2666 MT/s / DDR4, Intel® SSD DC S3700 800GB. Tencent InGame Purchase Machine Learning Platform*: OS: CentOS 7.3.1611. Testing by Intel May 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4, Intel® SSD DC S3700 800GB. NEW: 2S Intel® Xeon® Platinum 8180 processor, 2.5GHz, 28 cores, turbo and HT on, 192GB total memory, 12 slots / 16GB / 2666 MT/s / DDR4, Intel® SSD DC S3700 800GB. Kingsoft Cloud Image Processing and MySQL Cloud Service*: OS: CentOS 7.3.1611. Testing by Intel May 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4, S3700 800G SSD. NEW 1 (for general workload benchmarking): 2SIntel® Xeon® Platinum 8180 processor, 2.5GHz, 28 cores, turbo and HT on, 192GB total memory, 12 slots / 16GB / 2666 MT/s / DDR4, S3700 800G SSD. NEW 2 (for Intel® Optane™ SSD benchmarking): 2S Intel® Xeon® Platinum 8180 processor, 2.5GHz, 28 cores, turbo and HT on, 192GB total memory, 12 slots / 16GB / 2666 MT/s / DDR4, Intel® SSD DC P3700 2TB, Intel® Optane™ Solid State Drive DC P4800X 375GB. Qihoo Search Engine: CentOS Linux release 7.3.1611. Testing by Intel May 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4, SATA 1TB. NEW: 2S Intel® Xeon® Platinum 8180 processor, 2.5GHz, 28 cores, turbo and HT on, 192GB total memory, 12 slots / 16GB / 2666 MT/s / DDR4, Intel® SSD DC S3700 800GB. TongTech TongLink/Q middleware*: OS: CentOS 7.3.1611. Testing by Intel April 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, 128GB total memory, 8 slots / 16GB / 2133 MT/s / DDR4, Intel® SSD DC S3700 800GB. NEW: Intel® Xeon® Platinum 8180 processor, 2.5GHz, 28 cores, turbo and HT on, 192GB total memory, 12 slots / 16GB / 2666 MT/s / DDR4, Intel® SSD DC S3700 800GB. LeCloud Business Analytics*: OS: CentOS 7.3.1611 Linux kernel 4.9.8. Testing by Intel May 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4 RDIMM, Intel® SSD DC S3700 800GB. ICC version: v17.0.2 NEW: Intel® Xeon® Platinum 8180 processor, 2.5GHz, 28 cores, turbo and HT on, 192GB total memory, 12 slots / 16GB / 2666 MT/s / DDR4 RDIMM, Intel® SSD DC S3700 800GB. ICC version: v17.0.2 AsiaInfo Telco BSS*: AsiaInfo Telco BSS workload. OS: RHEL* 7.3. Testing by Intel & AsiaInfo May 2017. BASELINE: 4S Intel® Xeon® processor E7-8890 v4, 2.2GHz, 24 cores, turbo and HT on, 256GB total memory, 16 slots / 16GB / 1600 MT/s / DDR4, P3700 2T SSD x 2. NEW 1 (for general workload benchmarking and Intel® QAT benchmarking): 4S Intel® Xeon® Platinum 8180 processor, 2.5GHz, 28 cores, turbo and HT on, Intel® Chipset C627, 384GB total memory, 24 slots / 16GB / 2666 MT/s / DDR4, Intel® SSD DC P3700 2TB x 2. NEW 2 (for Intel® Optane™ SSD benchmarking): 4S Intel® Xeon® Platinum 8180 processor, 2.5GHz, 28 cores, turbo and HT on, Intel® Chipset C627, 384GB total memory, 24 slots / 16GB / 2666 MT/s / DDR4, Intel® SSD DC P3700 2TB x 2, Intel® Optane™ SSD DC P4800X 375G x 2.
- 141. 141 eBrisk eBLive A5000: OS: Windows Server* 2012 R2 Standard Build 9600. Test Environment: Bitrate is chosen as the maximum bitrate where the encoder is real-time at (up to 15Mbps for 4kp60/50, up to 10Mbps for 4kp30/25/24, up to 10Mbps for 1080p60/50). Original video was loaded to the RAM before the start of the encode. Test Clips ( https://media.xiph.org/video/derf/ ) : Amazon HDR UHD 24fps: downloaded then decoded using ffmpeg to yuv 420 10bit. Netflix Clips 2160p 60fps 10bit: Boat, Crosswalk, FoodMarket, Narrator, PierSeaside, RitualDance, Tango. SVT 1080p 50fps: crowd_run, ducks_take_off, old_town_cross, park_joy BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4 RDIMM, 360GB, Intel® SSD. NEW: Intel® Xeon® Platinum 8180 processor, 2.5GHz, 28 cores, turbo and HT on, 192GB total memory, 12 slots / 16GB / 2666 MT/s / DDR4 LRDIMM, 1 x 800GB, Intel® SSD DC S3710. Surf Communications Solutions Surf-HMP*: H.264 full HD video channels per core workload. BASELINE: 2S Intel® Xeon® processor E5-2690 v3, 2.6GHz, 24 cores, turbo and HT on, 32GB total memory, 4 DIMMs / 8GB / 2133 MT/s /DDR4, CentOS 7. NEW: 2S Intel® Xeon® processor Platinum 8168, 48 cores, turbo and HT on, 192GB total memory, 12 DIMMs/ 16GB / 2666 MT/s / DDR4 LRDIMM, 1 x 800GB, CentOS Linux 7. Beamr 5 (version 4.0) 4K HEVC encoding load, tested by Beamr June 2017. BASELINE: Data reflects performance measurements on 1 socket of a 2S Intel® Xeon® E5 2699 v4, 2.2GHz, 22 cores, turbo off / HT on, BIOS "Intel Corporation GRRFSDP1.86B.0271.R00.1510301446", 64 GB total memory, 8 slots / 8 GB / 2133 MT/s / DDR4 DIMM, 500GB Hard Disk, CentOS Linux* 7.2 kernel 3.10.0 NEW: Data reflects performance measurements on 1 socket of a 2S Intel® Xeon® Platinum 8180 processor, 2.5GHz, 28 cores, turbo off / HT on, BIOS "Intel Corporation SE5C620.86B.01.00.0412.020920172159", SMBIOS v2.8, 192 GB total memory, 12 slots / 16 GB / 2666 MT/s / DDR4 DIMM, 800GB Intel® SSD DC S3710, Ubuntu 16.04.2 LTS kernel 4.4.0-78-generic Sandvine Virtual Series OS: CentOS* Linux* release 7.3.1611 Kernal: Linux 3.10.0-514.6.2.el7.x86_64 Hypervisor: qemu-kvm-1.5.3-126.el7_3.3.x86_64 VNF sizing: 3vCPU (6 pCPU threads), 128 GB RAM Testing by Sandvine, June 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v3, 2.30 GHz, 18 cores, BIOS 2.0, 128 GB DDR4 2400Mhz (8 x 16 GB), 4x Intel® 82599ES 10 Gigabit Ethernet Controller NEW: 2S Intel® Xeon® Gold 6150 processor, 2.70 GHz, 18 cores, BIOS PLYDCRB1.86B.0128.R08.170324226, 128 GB DDR4 2400Mhz (8 x 16 GB), 4x Intel® 82599ES 10 Gigabit Ethernet Controller Keepixo workload. OS: CentOS Linux* 7.3 kernel 3.10.0. Testing by Keepixo May 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, BIOS 251.R01, 64GB total memory, 8 slots / 8GB / 2133 MT/s / DDR4 LRDIMM, CentOS Linux* 7.1 kernel 3.10.0. NEW: 2S Intel® Xeon® Platinum 8168 processor, 2.7GHz, 24 cores, turbo and HT on, BIOS 412, 192GB total memory, 12 slots / 16GB / 2600 MT/s / DDR4 LRDIMM, CentOS Linux* 7.3 kernel 3.10.0. Ericsson MediaFirst Video Processing UHD HEVC transcoding workload. OS: CentOS Linux* 7.2 kernel 3.10.0. Testing by Ericsson in May 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, BIOS 251.R01, 64GB total memory, 8 slots / 8GB / 2133 MT/s / DDR4 LRDIMM, CentOS Linux* 7.2 kernel 3.10.0. NEW: 2S Intel® Xeon® Platinum 8168 processor, 2.7GHz, 24 cores, turbo and HT on, BIOS 412, 192GB total memory, 12 slots / 16GB / 2600 MT/s / DDR4 LRDIMM, CentOS Linux* 7.2 kernel 3.10.0. 6WIND Turbo Ipsec* 1.4.2 OS: Ubuntu Linux* 16.04 kernel 4.4.0-77. Testing by 6WIND June 2017. BASELINE: 2S Intel® Xeon® processor E5-2680 v4, 2.1GHz, 14 cores, turbo and HT enabled (14c/28t), Dell R 530 BIOS 1.6.2, 64GB total memory, 4 slots / 16GB / 1600 MT/s / DDR4 LRDIMM, 2x Intel® Ethernet Connection XL710/X722, 1 x 500GB, Ubuntu 16.04 kernel 4.4.0-77-generic. NEW: 2S Intel® Xeon® Platinum 8170 processor, 2.1GHz, 26 cores, turbo and HT enabled (26c/52t), BIOS PLYDCRB1.86B.0131.R09.1704, 4GB total memory, 1 slot / 4GB / 2133 MT/s / DDR4 RDIMM, 2x Intel® Ethernet Connection XL710/X722, 1 x 500GB, Ubuntu 16.04 kernel 4.4.0-77-generic. Shenzhen TYDIC online charging system workload. OS: CentOS Linux release 7.2. Testing by Intel & TYDIC May 2017. BASELINE: 4S Intel® Xeon® processor E7-8890 V4, 2.2GHz, 24 cores, turbo and HT on, 256GB total memory, 16 slots / 16GB / 1600 MT/s / DDR4, Intel® SSD DC 3700 800GB. NEW: 2S Intel® Xeon® Platinum 8180 processor, 2.5GHz, 28 cores, turbo and HT on, 384GB total memory, 24 slots / 16GB / 2666 MT/s / DDR4, Intel® SSD DC 3700 800GB. MulticoreWare: x264 configuration: Version 0.150.2833 df79067, GCC 4.8.5, 8-bit profile, assembly until AVX-512 enabled, Videos: crowd_run, ducks_take_off, old_town_cross, park_joy 8-bit 4:2:0 1920x1080p @ 60fps, offline encoding with medium and veryslow presets averaged across 1.5, 4.0, 6.5, 9.0Mbps, 1080p60 live encoding with medium preset @ 6.5Mbps x265 configuration: Version 2.4+28-f850cdbe381c, GCC 4.8.5, 64-bit, main-10 profile, assembly until AVX2 enabled, Videos: Netflix Boat, Crosswalk, FoodMarket, PierSeaside, RitualDance, Tango 10-bit 4:2:0 4096x2160p @ 50fps, offline encoding with slower preset averaged across 12.5, 15.0, 17.5, 20.0Mbps, 1080p60 live encoding with ultrafast preset @ 6.5Mbps UHDkit configuration: Version 2.0+1-e7102605d241, GCC 4.8.5, 64-bit, main-10 profile, assembly until AVX2 enabled, Videos: CanalPlus_UHD_Astra_50fps, DFL_Supercup_UHD_50fps and Samsung_football_Barcelona vs Real Madrid 10-bit 4:2:2 3840x2160 @ 60fps, live encoding with 8 parallel x265 encoders (segmented per-GOP) @ 18Mbps BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, BIOS American Mega Trends v2.0, 256GB total memory, 16 slots / 16GB / 2133MHz / DDR4 2-rank DIMMs, 7 x 480GB, Intel® SSD DC S3510, CentOS Linux Release 7.2.1511 (Core), kernel 3.10.0 NEW: 2S Intel® Xeon® Platinum 8180 processor, 2.5GHz, 28 cores, turbo and HT on, BIOS Intel Corporation SE5C620.86B.01.00.0412.020920172159, 187GB total memory, 12 slots / 16GB / 2666MHz / DDR4 2-rank DIMM, 1x800GB Intel® SSD DC P3710, CentOS Linux Release 7.3.1611 (Core), kernel 3.10.0
- 142. 142 Pexip Infinity 2.6video conferencing handling maximum number of 1080HD video streams (withVP9 codec ) on a server. OS: Debian Linux* 8.8. Testing by Pexip June 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo on, HT on, 256GB DDR4 @2400 MHz, no storage was used by workload, Debian Linux 8.8. NEW: 2S Intel® Xeon® Platinum 8168 processor, 2.7GHz, 24 cores, turbo on, HT on, 192GB DDR4 @2666 MHz, no storage was used by workload, Debian Linux 8.8. Aerospike Database: The database was populated with 200 M records of 100 bytes each and benchmarked with the Aerospike Java Benchmark tool (https://github.com/aerospike/aerospike-client-java) . The workload simulated 95%/5% read/update ratio. Two Aerospike instances were launched on a single server forming a cluster. Each Aerospike instance was affinitized to a CPU socket and configured to use one of the 10GB NICs. Each 10GB NIC had its interrupt IRQs affinitized to a CPU socket. BASELINE (HSW-EP): Aerospike Server Enterprise 3.6.4 , CentOS 6.7, kernel version 2.6.32-573.3.1.el6.x86_64, 2 Intel® Xeon® processor E5-2697 v3, 2.6GHz, 28 cores, 128GB DDR4/1866, regular DIMM, 2x 10Gb network Intel X540-AT2 not bonded, no disk used – in memory workload, Clients: 8 client systems were used to concurrently submit queries to the servers and drive the workload. The same clients were used in both “baseline” and “new”. The clients were configured as follows: : E5- 2697 v3 128GB of memory and 10GB Intel X540-AT2 network The database was populated with 400 M records of 100 bytes each and benchmarked with the Aerospike Java Benchmark tool (https://github.com/aerospike/aerospike-client-java) . The workload simulated 95%/5% read/update ratio. Two Aerospike instances were launched on a single server forming a cluster. NEXT (BDW-EP, new software): Aerospike Server Enterprise 3.12.1, OS: CentOS 7.2 with kernel updated to 4.4.59, Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, BIOS SE5C610.86B.01.01.0016.033120161139, 128GB total memory, 16 DIMMs / 8GB / Configured Clock Speed: 1866 MHz / DDR4 DIMM, 2 x Intel® 82599ES 10 Gigabit Ethernet Controllers – all 4 ports on the 2 network controllers were bonded for an aggregate 40000Mb/s bond. No storage – in-memory workload NEW: Aerospike Server Enterprise 3.12.1, OS: CentOS 7.2 with kernel updated to 4.4.59, Intel® Xeon® Platinum processor 8180, 2.5GHz, 28 cores, turbo and HT on, BIOS SE5C620.86B.01.00.0412.020920172159 , 384GB total memory, 12 DIMMs / 32GB / Configured Clock Speed: 2666 MHz / DDR4 DIMM, 2 x Intel® 82599ES 10 Gigabit Ethernet Controllers – all 4 ports on the 2 network controllers were bonded for an aggregate 40000Mb/s bond. No storage – in-memory workload Clients: 8 client systems were used to concurrently submit queries to the servers and drive the workload. The same clients were used in both “baseline” and “new”. The clients were configured as follows: CentOS 7.2 with kernel 3.10.0-327. Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo and HT on, BIOS SE5C610.86B.01.01.0016.033120161139, 128GB total memory, 8 DIMMs / 16GB / Configured Clock Speed: 2400 MHz, 1 x Intel® 82599ES 10 Gigabit Ethernet Controllers EXASOL 6.0*: custom TPC-H like workload. OS: custom CentOS* 6.8. Testing by Intel and Exasol May 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2 GHz, 22 cores, turbo and HT on, BIOS 338.R00, 64GB total memory, 2133 MT/s / DDR4 LRDIMM, 1 TB SATA HDD, custom CentOS* 6.8 kernel 2.6.32-642. NEW: 2S Intel® Xeon® Platinum processor 8168, 2.7 GHz, 24 cores, turbo and HT on, BIOS XXXX, 192GB total memory, 12 slots / 16GB / 2666 MT/s / DDR4 LRDIMM, 1 TB SATA HDD, custom CentOS* 6.8 kernel 2.6.32-642. BASELINE: 4S Intel® Xeon® processor E7-8890 v4, 2.2 GHz, 24 cores, turbo and HT on, BIOS 338.R00, 256GB total memory, 32 slots / 8GB / 1600 MT/s / DDR4 LRDIMM, 300 GB SAS HDD + 1 TB SATA HDD, custom CentOS* 6.8 kernel 2.6.32-642. NEW 4: 4S Intel® Xeon® Platinum processor 8180, 2.5 GHz, 28 cores, turbo and HT on, BIOS XXXX, 768GB total memory, 24 slots / 32GB / 2666 MT/s / DDR4 LRDIMM, 800GB Intel® SSD DC S3700, custom CentOS* 6.8 kernel 2.6.32-642. The workload is not I/O and is not memory size bound. SAS Business Analytics*: SAS 9.4 m4 application running the 30 session SAS Mixed Analytics workload. OS: CentOS 7.2 kernel 3.10.0. Testing by Intel and SAS May 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo on, HT off, BIOS 275.R01.1603300531, 256GB total memory, 16 slots / 16GB / 2133 MT/s / DDR4 LRDIMM, 7 x 800GB Intel® SSD DC S3700, 1 x 2TB P3700 Intel® SSD Data Center Family for NVMe*, CentOS* 7.2 kernel 3.10.0. NEW: 2S Intel® Xeon® Platinum processor 8180, 2.5GHz, 28 cores, turbo on, HT off, BIOS 01.00.0412. 020920172159, 384GB total memory, 24 slots / 16GB / 2666 MT/s / DDR4 LRDIMM, 4 x 800GB, Intel SSD DC S3710 + 4 x 375GB P4800X Intel® Optane™ SSDs, CentOS* 7.2 kernel 3.10.0. IBM® DB2* (4 socket): DB2 v11.1.1.1. The IBM Big Data Insights Internal Heavy Multiuser Workload (BDInsights) is a multi-user data warehousing workload based on a retail environment. The workload is comprised with a mix of complex and intermediate queries. The scale factor for the workload is 3TB with 12 users. Testing by Intel® and IBM®. April/May 2017. BASELINE: 4S Intel® Xeon® processor E7-4890 v2, 2.8GHz, 15 cores, turbo on, HT on, BIOS 38.R02, 1.5TB total memory, 96 slots / 16GB / 1066 MT/s / DDR3 DIMM, IBM XIV Storage (132 10K RPM disks), SuSE Enterprise Linux* 11.3 kernel 3.0.101-0.47.71. NEXT GEN: 4S Intel® Xeon® processor E7-8890 v4, 2.2GHz, 24 cores, turbo on, HT on, BIOS 335.R00, 1.5TB total memory, 96 slots / 16GB / 1600 MT/s / DDR4 LRDIMM, 1 x 800GB, Intel® SSD DC S3700, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.16.1.el7.x86_64. NEW: 4S Intel® Xeon® Platinum processor 8180 v4, 2.5GHz, 28 cores, turbo on, HT on, BIOS 119.R05, 1.5TB total memory, 48 slots / 32GB / 2677 MT/s / DDR4 LRDIMM, 1 x 800GB, Intel® SSD DC S3700, Red Hat Enterprise Linux* X.X kernel 3.10.0-514.16.1.el7.x86_64.
- 143. 143 IBM® DB2*: DB2v11.1.1.1. The IBM Big Data Insights Internal Heavy Multiuser Workload (BDInsights) is a multi-user data warehousing workload based on a retail environment. The workload is comprised with a mix of complex and intermediate queries. The scale factor for the workload is 300GB with 12 users. Testing by Intel® and IBM®. April/May 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo, HT on, BIOS 0271.R00, 256GB total memory, 16 slots / 16GB / 2133 MT/s / DDR4 LRDIMM, 1 x 800GB Intel® SSD DC S3700, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.16.1.el7.x86_64. NEW: 2S Intel® Xeon® Platinum processor 8180, 2.5GHz, 28 cores, turbo on, HT on, BIOS 00.0412, 192GB total memory, 12 slots / 16GB / 2677 MT/s / DDR4 LRDIMM, 1 x 800GB Intel SSD DC S3700, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.16.1.el7.x86_64. IBM GSKit*: ICC 8.5, GSKit performance test. AES test: Block size 128K, 128 bit keys. RSA test: 2048 bit keys. SHA test: Block size 128K, SHA1 and SHA256 tested. Testing by IBM® and Intel® May 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, BIOS 0271.R00, 384GB total memory, 24 slots / 16GB / 1600 MT/s / DDR4 LRDIMM, 1 x 800GB Intel SSD DC S3700, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.16.1.el7.x86_64. NEW: 2S Intel® Xeon® Platinum processor 8180, 2.5GHz, 28 cores, turbo and HT on, BIOS 86B.01.00.0412.020920172159, 384GB total memory, 24 slots / 16GB / 2677 MT/s / DDR4 LRDIMM, 1 x 800GB Intel SSD DC S3700, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.16.1.el7.x86_64. IBM® dashDB Local*: dashDB Local v1.9. The IBM Big Data Insights Internal Heavy Multiuser Workload (BDInsights) is a multi-user data warehousing workload based on a retail environment. The workload is comprised with a mix of complex and intermediate queries. The scale factor for the workload is 300GB with 12 users. Testing by Intel® and IBM®. April/May 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo, HT on, BIOS 0271.R00, 256GB total memory, 16 slots / 16GB / 2133 MT/s / DDR4 LRDIMM, 1 x 800GB Intel SSD DC S3700, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.16.1.el7.x86_64. NEW: 2S Intel® Xeon® Platinum processor 8180, 2.5GHz, 28 cores, turbo on, HT on, BIOS 00.0412, 384GB total memory, 24 slots / 16GB / 2677 MT/s / DDR4 LRDIMM, 1 x 800GB Intel SSD DC S3700, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.16.1.el7.x86_64. IBM Algo Financial Modeler*: Workload is complex stochastic model with 1000 scenarios and valued over a 120 year period with 1 month projection steps. Testing by IBM April/May 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, BIOS 0271.R00, 192GB total memory, 24 slots / 8GB / 1600 MT/s / DDR4 LRDIMM, 1 x 800GB Intel® SSD DC S3700, Windows Server 2016 NEW: 2S Intel® Xeon® Platinum processor 8180, 2.5GHz, 28 cores, turbo and HT on, BIOS 86B.01.00.0412.020920172159, 192GB total memory, 12 slots / 16GB / 2677 MT/s / DDR4 LRDIMM, 1 x 800GB Intel SSD DC S3700, Windows Server* 2016. IBM Informix Warehouse Accelerator*: Analytical queries on a large set of sensor time series data, which originates from publicly available U.S. government data (NWIS - National Water Information System). The NoSQL data used in the workload represents periodic measurements (in 15 minute intervals) from more than 800 metering points over a period of roughly 100 years. This amounts to a total of more than 2.5 billion data records, each comprising the timestamp and multiple measurement values. The analytics workload performed on this data represents various aggregations typically done on time series data. Testing by IBM April 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, BIOS 0271.R00, 192GB total memory, 24 slots / 8GB / 1600 MT/s / DDR4 LRDIMM, 1 x 800GB Intel® SSD DC S3700, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.16.1.el7.x86_64. NEW: 2S Intel® Xeon® Platinum processor 8180, 2.5GHz, 28 cores, turbo and HT on, BIOS 86B.01.00.0412.020920172159, 192GB total memory, 12 slots / 16GB / 2677 MT/s / DDR4 LRDIMM, 1 x 800GB Intel SSD DC S3700, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.16.1.el7.x86_64. Compuverde vNAS workload. OS: Centos 7 kernel 3.9. Testing by Intel and Compuverde June 2017. 3 Compuverde storage nodes: Intel® Xeon® Platinum 8180 Processor, 2.5GHz, 28 cores, turbo and HT on, 96GB total memory, 12x 8GB DDR4-2666, 2 x 400GB Intel SSD DC P3700 + 10 x 900GB SAS 10K (HBA), 2 x Intel X520 10GbE NIC using custom Centos 7 installation w/ kernel 3.9. Zebware ZebEC: Payload: 8KiB, Redundancy configuration : 4+2, Erasures: 1, Setting: Performance, OS: Ubuntu Linux* 16.04 kernel 4.4. Testing by Zebware May 2017. BASELINE: 2S Intel® Xeon® processor E5-2680 v4, 2.4GHz, 14 cores, turbo and HT on, 256GB total memory, DDR4 DIMM, Intel SSD. NEW: 2S Intel® Xeon® Platinum processor 8168 v4, 2.7GHz, 24 cores, turbo and HT on, 192 GB total memory, Intel SSD. QuikFynd Enterprise Search*: version 2.7.4 workload. OS: Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.x. Testing by Intel and QuikFynd May 2017. BASELINE: 2S Intel® Xeon® processor E5-2697 v3 @ 2.6GHz, 28 cores, turbo and HT on, 8x16GB 2400MHz DDR4, BIOS 86B0271.R00, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.x. NEW: 2S Intel® Xeon® Platinum 8180 processor @ 2.5GHz, 56 cores, turbo and HT on, 12x16GB 2666MHz DDR4, BIOS SE5C620.86B.01.00.0412.020920172159, Red Hat Enterprise Linux* 7.3 kernel 3.10.0- 514.x.
- 144. 144 InterSystems Caché. InterSystemsTrakCare* 2017. OS: Red Hat Enterprise Linux* 7.3. Testing by Intel and InterSystems May 2017. BASELINE: 4S Intel® Xeon® processor E7-8890 v3, 2.5GHz, 72 cores, turbo and HT on, 2048GB total memory, 4 x 2TB Intel® SSD Data Center P3700 Family for NVMe*, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-327.28.2.el7.x86_64. NEW: 4S Intel® Xeon® Platinum processor 8180, 2.5GHz, 112 cores, turbo and HT on, 384GB total memory, 3 x 2TB Intel SSD Data Center P3700 Family for NVMe, Red Hat Enterprise Linux* 7.3 kernel 3.10.0- 327.28.2.el7.x86_64. InterSystems Caché. InterSystems TrakCare* 2017. OS: Red Hat Enterprise Linux* 7.3. Testing by Intel and InterSystems May 2017. BASELINE: 2S Intel® Xeon® processor E5-2680 v3, 2.5GHz, 24 cores, turbo and HT on, 512GB total memory, 4 x 2TB Intel® SSD Data Center P3700 Family for NVMe*, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.el7.x86_64. NEW: 2S Intel® Xeon® Platinum processor 8168, 2.5GHz, 48 cores, turbo and HT on, 192GB total memory, 3 x 2TB Intel SSD Data Center P3700 Family for NVMe, Red Hat Enterprise Linux* 7.3 kernel 3.10.0- 327.28.2.el7.x86_64. Altibase 7.1* : Altibase 7.1* workload. OS: CentOS* 7.3 kernel 3.10.0-514. Testing by Intel and Altibase May 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, BIOS SE5C610.86B.01.01.0014.121820151719, 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4 LRDIMM, 1TB HDD(SATA 7200rpm), CentOS 7.2 kernel 3.10.0-327. NEW: 2S Intel® Xeon® Platinum processor 8180, 2.5GHz, 28 cores, turbo and HT on, BIOS SEC5C620.86B.01.00.0265.083120160756, 192GB total memory, 12 slots / 16GB / 2666 MT/s / DDR4 LRDIMM, 800GB SSD (Intel® SSD DC S3710), CentOS 7.3 kernel 3.10.0-514. Intel Compiler 17.0.2 BASELINE: 4S Intel® Xeon® processor E7-8890 v4, 2.2GHz, 24 cores, turbo and HT on, BIOS BRBDXD1.86B.0338.R00, 512GB total memory, 32 slots / 16GB / 1600 MT/s / DDR4 LRDIMM, 800GB Intel® S3700 SSD DC 2EA in RAID 0, CentOS 7.2 kernel 3.10.0-327. NEW: 4S Intel® Xeon® Platinum processor 8180, 2.5GHz, 28 cores, turbo and HT on, BIOS PLYDCRB1.86B.0087.D08.160524155, 768GB total memory, 24 slots / 32GB / 2666 MT/s / DDR4 LRDIMM, 1.2TB SSD(Intel® SSD DC S3710), CentOS 7.3 kernel 3.10.0-514. Intel Compiler 17.0.2 SUNJESOFT Goldilocks: Goldilocks workload. OS: CentOS 7.3 kernel 3.10.0. Testing by Intel and SUNJESOFT April 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, BIOS SE5C610.86B.01.01.0014.121820151719, 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4 LRDIMM, 1TB HDD(SATA 7200rpm), CentOS 7.2 kernel 3.10.0-327. NEW: 2S Intel® Xeon® Platinum processor 8168, 2.7GHz, 24 cores, turbo and HT on, BIOS SEC5C620.86B.01.00.0265.083120160756, 192GB total memory, 12 slots / 16GB / 2666 MT/s / DDR4 LRDIMM, 800GB SSD(Intel SSD DC S3710), CentOS 7.3 kernel 3.10.0-327. Intel Compiler 17.0.2 BASELINE: 4S Intel® Xeon® processor E7-8880 v4, 2.2GHz, 22 cores, turbo and HT on, BIOS BRBDXD1.86B.0338.R00, 256GB total memory, 16 slots / 16GB / 1600 MT/s / DDR4 LRDIMM, 800GB SSD(Intel SSD DC S3700, CentOS 7.2 kernel 3.10.0-327. NEW: 4S Intel® Xeon® Platinum processor 8164, 2.0GHz, 26 cores, turbo and HT on, BIOS PLYDCRB1.86B.0087.D08.160524155, 384GB total memory, 24 slots / 16GB / 2666 MT/s / DDR4 LRDIMM, 1.2TB SSD(Intel SSD DC S3710), CentOS 7.3 kernel 3.10.0-514. Intel Compiler 17.0.2 Innovatrics fingerprint identification benchmark, 14 million fingerprints DB loaded in RAM. OS: CentOS Linux release 7.3.1611 kernel 3.10.0-327.el7.x86_64. Testing by Intel and Innovatrics May-June 2017. Performance measured with two instance of the application binding each instance to each NUMA node . BASELINE: 2S Intel® Xeon® processor E5-2699 v3, 2.3GHz, 18 cores, turbo and HT on, 128GB total memory (2DIMM per channel ) 16 slots / 8GB / 2133 MT/s / DDR4 RDIMM NEXT GEN: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, 128GB total memory, (1DIMM per channel ) 8 slots / 16GB / 2400 MT/s / DDR4 RDIMM NEW: 2S Intel® Xeon® Platinum processor 8180 , 2.5GHz, 28 cores, turbo and HT on, 192GB total memory, (1DIMM per channel ) 12 slots / 16GB / 2666 MT/s / DDR4 RDIMM Software AG Apama* Streaming Analytics processing workload. OS: CentOS 7.2*. These results are all generated with the source of data on a 2 socket Xeon Platinum 8180 system, which sent events over a pair of bonded 10GbE NICs to the 4 socket system under test. The throughput is in millions of events (location updates) per second. Testing by Software AG June 2017. BASELINE: 4S Intel® Xeon® processor Intel® Xeon® E7-8890 v4, 2.2 GHz, BRBDXSD1.86B.0338.R00.1603162127, 512 GB memory, 32 * 16GB DDR4 DIMMs, 1866MHz NEW: 4S Intel® Xeon® Platinum 8180 processor, 2.5 GHz, PLYDCRB1.86B.0119.R05.1701181843, 1.5TB memory, 24 * 64GB DDR4 DIMMs, 2666 MHz Nari Grid Dispatching System (GDS): OS: RHEL 7.2. Testing by Intel May 2017. Performance metric: response time(s) BASELINE: 4S Intel® Xeon® processor E7-8890 v4, 2.2GHz, 24 cores, turbo and HT on, 256GB total memory, 16 slots / 16GB / 1600 MT/s / DDR4, P3700 2T SSD x 2. NEW (for general workload benchmarking): Intel® Xeon® Platinum processor 8180, 2.5GHz, 28 cores, turbo and HT on, Lewisburg C627, 384GB total memory, 24 slots / 16GB / 2666 MT/s / DDR4, P3700 2T SSD x 2.
- 145. 145 NRI* Business OrientedSolution*: Real-time stock calculation; Benchmark schema derived from TPC-H* queries. OS: Red Hat Enterprise Linux* 7.3 kernel 3.10-514. Testing by NRI Apr 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo on and HT off, BIOS 275.R01 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4 LRDIMM, 1 x 240GB, Intel SSD DC S3500, Red Hat Enterprise Linux* 7.3 kernel 3.10-514. NEW: Intel® Xeon® Platinum processor 8180, 2.5GHz, 28 cores, turbo on and HT off, BIOS 412, 192GB total memory, 12 slots / 16GB / 2666 MT/s / DDR4 LRDIMM, 1 x 800GB, Intel SSD DC S3710, Red Hat Enterprise Linux* 7.3 kernel 3.10-514. Yonyou Enterprise Resource Planning Platform: OS: RHEL 7.3. Testing by Intel May 2017. BASELINE: 4S Intel® Xeon® processor E7-8890 v4, 2.2GHz, 24 cores, turbo and HT on, 256GB total memory, 16 slots / 16GB / 1600 MT/s / DDR4, P3700 2T SSD x 2. NEW 1 (for general workload benchmarking): 4S Intel® Xeon® Platinum processor 8180, 2.5GHz, 28 cores, turbo and HT on, Intel® Chipset C627, 384GB total memory, 24 slots / 16GB / 2666 MT/s / DDR4, P3700 2T SSD x 2. NEW 2 (for Intel Optane benchmarking): 4S Intel® Xeon® Platinum processor 8180, 2.5GHz, 28 cores, turbo and HT on, Intel® Chipset C627, 384GB total memory, 24 slots / 16GB / 2666 MT/s / DDR4, P3700 2T SSD x 2, Intel® Optane™ SSD DC P4800X x 2. Kingdom STS: OS: Windows Server 2012 R2 (x64). Testing by Intel May 2017. BASELINE: 4S Intel® Xeon® processor E7-8890 v4, 2.2GHz, 24 cores, turbo and HT on, 256GB total memory, 16 slots / 16GB / 1600 MT/s / DDR4, Intel® SSD DC P3700 2TB. NEW: 4S Intel® Xeon® Platinum 8180 processor, 2.5GHz, 28 cores, turbo and HT on, 384GB total memory, 24 slots / 16GB / 2666 MT/s / DDR4, Intel® SSD DC P3700 2TB. QlikView*: HWE workload. OS: Windows 2016 Server*. Testing by Intel and Qlik May 2017. BASELINE: 2S Intel® Xeon® E5-2699 v3, 2.3GHz, 18 cores, turbo and HT on, BIOS 63.R00, 64GB total memory, 1600 MT/s / DDR4 LRDIMM, 1TB WDC SATA HDD. NEXT GEN: 2S Intel® Xeon® E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, BIOS 63.R00, 64GB total memory, 1600 MT/s / DDR4 LRDIMM, 1TB WDC SATA HDD. NEW: 2S Intel® Xeon® Platinum processor 8168, 2.7 GHz, 24 cores, turbo and HT on,192GB total memory, 12 slots / 16GB / 2666 MT/s / DDR4 LRDIMM, 1 TB SATA HDD. The workload is not I/O bound and is not memory size bound. Oracle Database 12c R2: BASELINE: 1 – Node, 2 x Intel® Xeon® Processor E5-2699 v4, 8 x 32 GB DDR4 – 2134 DIMM, 1DPC, <256 > GB Total Memory on <OS Red Hat Enterprise Linux* 6.5 kernel 2.6.32-431.e16.x86_64>. Benchmark: < Star Schema Benchmark> NEW: 1 – Node, 2 x Intel® Xeon® Platinum 8180 processor @ 2.70 GHz ( 28 core parts used for experiment ) < Platform : Wolfpass> with 12 x 16 GB DDR4 – 2400 DIMM, 1DPC < 192 > GB Total Memory on <OS Linux 4.11.0 x86_64>, Benchmark : < Star Schema Benchmark> Oracle Database In-Memory 12c R2: BASELINE: 1 – Node, 2 x Intel® Xeon® Platinum 8180 processor, Intel® AVX2 @ 2.70 GHz ( 28 core parts used for experiment ) < Platform : Wolfpass> with 12 x 16 GB DDR4 – 2400 DIMM, 1DPC < 192 > GB Total Memory on <OS Linux 4.11.0 x86_64>, Benchmark : < Star Schema Benchmark> NEW: 1 – Node, 2 x Intel® Xeon® Platinum 8180 processor, Intel® AVX-512 @ 2.70 GHz ( 28 core parts used for experiment ) < Platform : Wolfpass> with 12 x 16 GB DDR4 – 2400 DIMM, 1DPC < 192 > GB Total Memory on <OS Linux 4.11.0 x86_64>, Benchmark : < Star Schema Benchmark> Oracle TimesTen 11.2.2.8.20: Memory Interleaved, 16 rows / hashpage BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores per socket, BIOS: Intel GRRFSDP1.86B.0271.R00.1510301446 (Release Date: 10/30/2015), 8x32GB, DDR4-2133 (Samsung* M386A4G40DM0-CPB), OS: RHEL 6.5 2.6.32-431.23.3.el6.x86_64, Intel® SSD DC P3600 NVMe* card NEW: 2S Intel® Xeon® Platinum 8180 processor, 2.7GHz, 24 cores per socket, BIOS: Intel SE5C620.86B.01.00.0412.020920172159 (Release Date: 02/09/2017), 8x16GB, DDR4-2666 (Micron* 18ASF2G72PDZ- 2G6B1), OS: RHEL 7.3 3.10.0-514.el7.x86_64, Intel® SSD DC P3600 NVMe* card Altair RADIOSS*: RADIOSS 2017, Neon 1M 8ms benchmark workload. NEON front car crash refined model with 1 million of elements, first 8ms run. BASELINE: Altair RADIOSS 14 on Red Hat Enterprise Linux* 6.5, 2 Intel® Xeon® processor E5-2697 v3, 2.6GHz, 28 cores, Hyper-Threading Technology with 28 MPI x 2 OpenMP, 64GB DDR3/1833, regular DIMM, Intel® SSD DC S3700 800GB, 1Gb network, Source is Altair internal as of April 1, 2016. Next GEN: Altair RADIOSS 2017 on Red Hat Enterprise Linux 6.5, 2 Intel® Xeon® processor E5-2699 v4, 2.2GHz, 44 cores, Hyper-Threading Technology with 44 MPI x 2 OpenMP, 64GB DDR3/1833, regular DIMM, Intel® SSD DC S3700 800GB, 1Gb network, Bios SE5C620.86B.01.00.0412, Source is Altair internal as of April 1, 2017. NEW: Altair RADIOSS 2017 on CentOS Linux 7.2, 2 Intel® Xeon® Gold 6148 processor, 2.4GHz, 40 cores, Hyper-Threading Technology with 40 MPI x 2 OpenMP, 192GB DDR4/2666, regular DIMM, Intel® SSD DC S3700 800GB, 1Gb network, Bios 0271.R00, Source is Altair internal as of April 11, 2017.
- 146. 146 CST EM Studio,time domain solver standard benchmark. Baseline: 2S Intel® Xeon® processor E5-2697 v3, 2.6GHz, HT on, 14 cores, turbo on, 64GB DDR4, CentOS 6.4 Generation1: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, HT on, 18 cores, turbo on, BIOS GRRFSDP1.86B.0275.R01.1603300531, 64GB total memory, CentOS Linux* 6.8 kernel 2.6.32. Generation2: 2S Intel® Xeon® Gold 6148 processor, 2.4GHz, HT on, 20 cores, turbo on, BIOS SE5C620.86B.01.00.0412.020920172159, 192GB total memory, CentOS Linux* 7.3.1611 kernel 3.10.0 SIMULIA Abaqus 2017x standard benchmarks tested by Intel, March 2017 BASELINE: 2S Intel® Xeon® processor E5-2698 v3, 2.3GHz, 16 cores, turbo and HT on, BIOS SE5C610.86B.01.01.5008.081020151115, 128GB total memory, 8 memory channels / 8x16GB / 2133 MT/s / DDR4, Red Hat Enterprise Linux* 7.3, kernel 3.10.0-229.20.1.el6.x86_64.knl2, Intel® SSD S3700 Series. NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo and HT on, BIOS GRRFSDP1.86B0271.R00.1510301446, 128GB total memory, 8 memory channels / 8x16GB / 2400 MT/s / DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327.el7.x86_64, 2x Intel® SSD S3500 Series RAID0 array. NEW: 2S Intel® Xeon® Gold processor 6148, 2.4GHz, 20 cores, turbo boost tech on, turbo energy efficient off, HT on, BIOS SE5C620.86B.01.00.0412.020920172159, 192GB total memory, 12 memory channels / 12x16GB / 2666 MT/s / DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327.el7.x86_64, 2x Intel® SSD S3500 Series RAID0 array. Note*: MKL shipped with 2017x Abaqus release is swapped with MKL 2017u1, and added environment variable: MKL_ENABLE_INSTRUCTIONS=AVX512 [Not an officially supported software kit] SIMULIA Abaqus 2017x Explicit benchmarks tested by Intel, April 2017 BASELINE: 2S Intel® Xeon® processor E5-2697 v3, 2.6GHz, 14 cores, turbo and HT on, 128GB total memory, 8 memory channels / 8x16GB / 2133 MT/s / DDR4, Red Hat Enterprise Linux* 7.3, kernel 3.10.0- 514.6.2.0.1.el7.x86_64.knl1, Intel® SSD S3700 Series. NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo and HT on, BIOS GRRFSDP1.86B0271.R00.1510301446, 128GB total memory, 8 memory channels / 8x16GB / 2400 MT/s / DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327.el7.x86_64, 2x Intel® SSD S3500 Series RAID0 array. NEW: 2S Intel® Xeon® Gold processor 6148, 2.4GHz, 20 cores, turbo boost tech on, turbo energy efficient off, HT on, BIOS SE5C620.86B.01.00.0412.020920172159, 192GB total memory, 12 memory channels / 12x16GB / 2666 MT/s / DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327.el7.x86_64, 2x Intel® SSD S3500 Series RAID0 array. LSTC LS-DYNA Baseline: 2S Intel® Xeon® processor E5-2697 v3, 2.3GHz, 14 cores, turbo and HT on, BIOS SE5C610.86B.01.01.5008.081020151115, 128GB total memory, 8 memory channels / 8x16GB / 2133 MT/s / DDR4, Red Hat Enterprise Linux* 7.3, kernel 3.10.0-229.20.1.el6.x86_64 Next Gen: Dual Socket Intel® Xeon® processor E5-2697 v4 2.3 GHz, Turbo mode ON , 18 Cores/Socket, 36 Cores (HT off), DDR4 128GB, 2400 MHz, Wildcat Pass Platform. Disk: 800GB Intel SSD Kernel: 3.10.0- 229.20.1.el6.x86_64 New: Dual Socket Intel® Xeon® Gold 6148 processor 1.8 GHz, Turbo mode ON , 28 Cores/Socket, 56 Cores (HT off), DDR4 128GB, 2667 MHz, Wolf Pass Platform. Disk: 800GB Intel SSD Kernel:3.10.0- 327.e17.x86-64 INTES PERMAS: Baseline: 2S Intel® Xeon® processor E5-2697 v3, 2.6GHz, 14 cores, turbo on, HT off, NUMA on, BIOS 46.R00, 256 GB total memory (8x 32GB w/ 2400 MT/s, DDR4 LRDIMM), 4x Intel® SSD DC P3600 1.6 TB in RAID 0 (stripe size 64k). CentOS* Linux* 6.6, kernel 4.1.2. Intel® Composer 2015.2.164. INTES PERMAS V15.10.10. Next generation: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo on, HT off, NUMA on, BIOS 338.R00, 256 GB total memory (8x 32GB w/ 2400 MT/s, DDR4 LRDIMM), 4x Intel® SSD DC P3600 2 TB in RAID 0 (stripe size 64k). CentOS Linux* release 7.2, kernel 3.10.0-327.13.1.el7.x86_64. Intel® Composer 2015.5.223. INTES PERMAS V16.00. New: 2S Intel® Xeon® Gold processor 6148, 2.4 GHz, 20 cores, turbo on, HT off, NUMA on, BIOS SE5C620.86B.01.00.0412.020920172159, 384 GB total memory (12x 32GB w/ 2400 MT/s, DDR4 LRDIMM), 3x Intel® SSD DC P3600 2 TB in RAID 0 (stripe size 64k), CentOS* Linux* release 7.3, kernel 3.10.0-514.10.2.el7.x86_64. Intel® Composer 2015.7.235. INTES PERMAS V16.00. INTES PERMAS (2 socket): E5-2680: 2S Intel® Xeon® processor E5-2680, 2.7 GHz, 8 cores, turbo on, HT off, NUMA on, 160 GB total memory (8x 16 GB & 8x 4 GB w/ 1333 MT/s, DDR3), 2x SSD OCZ Vertex 2 in RAID 0 (stripe size 64k) and 6x Seagate ST3300657SS (stripe size 256k). CentOS* Linux* 7.1, kernel 3.10.0-229.20.1, Intel® Composer 2015.2.164. INTES PERMAS V15.10.12. E5-2697v2: 2S Intel® Xeon® processor E5-2697 v2, 2.7 GHz, 12 cores, turbo on, HT off, NUMA on, BIOS 28D690, 128 GB total memory (8x 16 GB w/ 1600 MT/s, DDR3), 4x SSD Samsung 840pro in RAID0 (stripe size 64k), CentOS* Linux* 6.6, kernel 2.6.32-573.7.1, Intel® Composer 2011_sp1.13.367. INTES PERMAS V15.10.12. E5-2697v3: 2S Intel® Xeon® processor E5-2697 v3, 2.6GHz, 14 cores, turbo on, HT off, NUMA on, BIOS 46.R00, 256 GB total memory (8x 32GB w/ 2400 MT/s, DDR4 LRDIMM), 4x Intel® SSD DC P3600 2 TB in RAID 0 (stripe size 64k). CentOS* Linux* 6.6, kernel 4.1.2. Intel® Composer 2015.2.164. INTES PERMAS V15.10.10. E5-2697v4: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo on, HT off, NUMA on, BIOS 338.R00, 256 GB total memory (8x 32GB w/ 2400 MT/s, DDR4 LRDIMM), 4x Intel® SSD DC P3600 2 TB in RAID 0 (stripe size 64k). CentOS Linux* release 7.2, kernel 3.10.0-327.13.1.el7.x86_64. Intel® Composer 2015.5.223. INTES PERMAS V16.00. 6148: 2S Intel® Xeon® Gold processor 6148, 2.4 GHz, 20 cores, turbo on, HT off, NUMA on, BIOS SE5C620.86B.01.00.0412.020920172159, 384 GB total memory (12x 32GB w/ 2400 MT/s, DDR4 LRDIMM), 3x Intel® SSD DC P3600 1.6 TB in RAID 0 (stripe size 64k), CentOS* Linux* release 7.3, kernel 3.10.0-514.10.2.el7.x86_64. Intel® Composer 2015.7.235. INTES PERMAS V16.00.
- 147. 147 INTES PERMAS (4socket): E7-8890v3: 4S Intel® Xeon® processor E7-8890 v3, 2.5 GHz, 18 cores, turbo on, HT off, NUMA on, 256 GB total memory (16x 16 GB w/ 1333 MT/s, DDR3), 4x Samsung 840pro 512GB in RAID 0 (stripe size 64k) and 6x Seagate ST3300657SS (stripe size 256k). CentOS* Linux* 7.1, kernel 3.10.0-229.20.1, Intel® Composer 2015.2.164. INTES PERMAS V15.10.12. E7-8890v4: 4S Intel® Xeon® processor E7-8890 v4, 2.2 GHz, 24 cores, turbo on, HT off, NUMA on, BIOS 338.R00, 256 GB total memory (16x 16 GB w/ 1333 MT/s, DDR3), 4x Intel® SSD DC P3600 1.6 TB in RAID 0 (stripe size 64k), CentOS* Linux* 7.2, kernel 3.10.0-327.13.1, Intel® Composer 2015.5.223. INTES PERMAS V15.99.015. E7-8890v4 (4TB): 4S Intel® Xeon® processor E7-8890 v4, 2.2 GHz, 24 cores, turbo on, HT off, NUMA off, BIOS 338.R00, 4096 GB total memory (32x 128 GB w/ 1867 MT/s, DDR4), no I/O, in-memory run. CentOS* Linux* 7.2, kernel 3.10.0-514.2.2.el7.x86_64, Intel® Composer 2015.7.235. INTES PERMAS V16.00.223. 8180: 4S Intel® Xeon® processor Platinum 8180, 2.5 GHz, 28 cores, turbo on, HT off, NUMA off, BIOS 338.R00, 768 GB total memory (24x 32GB w/ 2400 MT/s, DDR4 LRDIMM), 4x Intel® SSD DC P3600 1.6 TB in RAID 0 (stripe size 64k). CentOS Linux* release 7.3, kernel 4.11.0-1.el7.x86_64. Intel® Parallel Studio 2017.2.174 INTES PERMAS V16.00.302 MSC Nastran V2017 standard benchmarks tested by Intel, June, 2017 BASELINE: 2S Intel® Xeon® processor E5-2680 v3, 2.5GHz, 12 cores, turbo and HT on, BIOS SE5C610.86B.01.01.0016.033120161139, 128GB total memory, 8 memory channels / 8x16GB / 2333 MT/s / DDR4, Red Hat Enterprise Linux* 7.1 kernel 2.6.32-431.el6.x86_64. single 800GB Intel® SSD S3700 NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo and HT on, BIOS SE5C610.86B.01.01.0016.033120161139, 128GB total memory, 8 memory channels / 8x16GB / 2400 MT/s / DDR4, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-229.20.1.el6.x86_64.knl2. single 800GB Intel® SSD S3700 NEW: 2S Intel® Xeon® Gold processor 6148, 2.4GHz, 20 cores, turbo and HT on, BIOS version 412, 256GB total memory, 12 memory channels / 16x16GB / 2666 MT/s / DDR4, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.el7.x86_64. single 800GB Intel® SSD S3700 Siemens STAR-CCM+* 12.05.023: Lemans 17M workload. Testing by Intel May 2017. BASELINE: 2S Intel® Xeon® processor E5-2697 v3, 2.6GHz, 14 cores, turbo and HT on, CoD on, BIOS 46.R00, 64GB total memory, 8 slots / 8GB / 2133 MT/s / DDR4 RDIMM, 800GB Intel® SSD DC S3700, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.el7.x86_64. NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4, 2.2GHz, 18 cores, turbo and HT on, CoD off, BIOS 271.R00, 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4 RDIMM, 800GB Intel SSD DC S3700, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.el7.x86_64. NEW: 2S Intel® Xeon® Gold 8168 processor, 2.4GHz, 20 cores, turbo and HT on, BIOS 01.00.0412, 192GB total memory, 12 slots / 16GB / 2666 MT/s / DDR4 RDIMM, 800GB Intel SSD DC S3700, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.el7.x86_64. ANSYS® Fluent 18.1 standard benchmarks tested by Intel, March 2017 BASELINE: 2S Intel® Xeon® processor E5-2698 v3, 2.3GHz, 16 cores, turbo and HT on, BIOS SE5C610.86B.01.01.5008.081020151115, 128GB total memory, 8 memory channels / 8x16GB / 2133 MT/s DDR4, Red Hat Enterprise Linux* 7.3, kernel 3.10.0-229.20.1.el6.x86_64.knl2. Turbo enabled, Hyper-threading enabled, Performance power settings. NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo and HT on, BIOS SE5C610.86B.01.01.0016.033120161139, 128GB total memory, 8 memory channels / 8x16GB / 2400 MT/s DDR4, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-229.20.1.el6.x86_64.knl2. Turbo enabled, Hyper-threading enabled, Performance power settings. NEW: 2S Intel® Xeon® Gold processor 6148, 2.4GHz, 20 cores, turbo and HT on, BIOS version 412, 192GB total memory, 12 memory channels / 12x16GB / 2666 MT/s DDR4, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.el7.x86_64. 2666 MT/s DDR4 RDIMMs. Turbo enabled, Hyper-threading enabled, Performance power settings, Energy Efficient Turbo Disabled. ANSYS® Mechanical18.1 standard benchmarks tested by Intel, March 2017. In-core solver used throughout. BASELINE: 2S Intel® Xeon® processor E5-2698 v3, 2.3GHz, 16 cores, turbo and HT on, BIOS SE5C610.86B.01.01.5008.081020151115, 128GB total memory, 8 memory channels / 8x16GB / 2133 MT/s DDR4, Red Hat Enterprise Linux* 7.3, kernel 3.10.0-229.20.1.el6.x86_64.knl2. Turbo enabled, Hyper-threading enabled, Performance power settings. 1x 800 GB SSD Intel ® SSD DC S3500 Series for I/O NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo and HT on, BIOS SE5C610.86B.01.01.0016.033120161139, 128GB total memory, 8 memory channels / 8x16GB / 2400 MT/s DDR4, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-229.20.1.el6.x86_64.knl2. Turbo enabled, Hyper-threading enabled, Performance power settings. 1x 800 GB SSD Intel SSD DC S3500 Series for I/O NEW: 2S Intel® Xeon® Gold processor 6148, 2.4GHz, 20 cores, turbo and HT on, BIOS version 412, 192GB total memory, 12 memory channels / 12x16GB / 2666 MT/s DDR4, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.el7.x86_64. 2666 MT/s DDR4 RDIMMs. Turbo enabled, Hyper-threading enabled, Performance power settings, Energy Efficient Turbo Disabled. 1x 800 GB SSD Intel SSDSC2BA80 for I/O. ANSYS® CFX: Internal benchmark: Solver wall clock time for 3 Ansys’ standard benchmarks (perf_Airfoil_10M_R14, perf_LeMansCar, perf_Pump). 2x2: CFX 17 on Xeon® E5-2697 v4 compared to CFX 18 on Xeon® Gold 6148. Testing by Intel and ANSYS May 2017. BASELINE: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo on, HT off, 128GB total memory @2400 MHz, no storage used by benchmarks, Red Hat* Enterprise Linux Server release 6.4. NEW: 2S Intel® Xeon® Gold processor 6148, 2.4GHz, 20 cores, turbo on, HT off, 192GB total memory @2666MHz, no storage used by benchmarks, CentOS Linux* 7.2. IHS Markit Analytics Risk Engine*: internal synthetic portfolio. OS: Windows Server* 2016. Testing by Intel and IHS Markit May 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.3GHz, 22 cores, turbo on, HT on, 128GB DDR4 @2400 MHz, no storage was used by workload, Windows server 2012 R2. NEW: 2S Intel® Xeon® Platinum 8168 processor, 2.7GHz, 24 cores, turbo on, HT on, 192GB DDR4 @2666 MHz, no storage was used by workload, Windows Server 2016.
- 148. 148 Siemens NX.Nastran V11standard benchmarks tested by Intel, June, 2017 BASELINE: 2S Intel® Xeon® processor E5-2680 v3, 2.5GHz, 12 cores, turbo and HT on, BIOS SE5C610.86B.01.01.0016.033120161139, 128GB total memory, 8 memory channels / 8x16GB / 2333 MT/s / DDR4, Red Hat Enterprise Linux* 7.1 kernel 2.6.32-431.el6.x86_64. single 800GB Intel® SSD S3700 NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo and HT on, BIOS SE5C610.86B.01.01.0016.033120161139, 128GB total memory, 8 memory channels / 8x16GB / 2400 MT/s / DDR4, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-229.20.1.el6.x86_64.knl2. single 800GB Intel® SSD S3700 NEW: 2S Intel® Xeon® Gold processor 6148, 2.4GHz, 20 cores, turbo and HT on, BIOS version 412, 256GB total memory, 12 memory channels / 16x16GB / 2666 MT/s / DDR4, Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.el7.x86_64. 800GB Intel® SSD S3700 ESI Group VPS 2016.01, ifort 16 update 2, impi 5.1.3, multi-core build, explicit models as input BASELINE: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 36 cores, turbo and HT on, BIOS SE5C610.86B.01.01.0016.033120161139, 128GB total memory (16 x 8GB w/ 2133 MT/s, DDR4 LRDIMM), Red Hat Enterprise Linux* Server release 7.3, kernel 3.10.0-514.6.2.0.1.el7.x86_64 NEW: 2S Intel® Xeon® Gold 6148 processor, 2.4GHz, 40 cores, turbo and HT on, BIOS SE5C620.86B.01.00.0412.020920172159, 192GB total memory (12 x 16GB w/ 2666 MT/s, DDR4 LRDIMM), Red Hat Enterprise Linux* 7.3, kernel 3.10.0-514.6.2.0.1.el7.x86_64 Pixar’s RenderMan*: Version 21 July 2016. Commercial release binary built with –axSSE3. Tests performed in March 2017. Workloads: PxrPathTracer.rib, PxrUPBP.rib (both bundled with commercial released version). BASELINE: Executed with 72 threads. 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 36 cores, turbo and HT on, BIOS 86B0271.R00, 8x16GB 2400MHz DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. NEW: Executed with 80 threads. 2S Intel® Xeon® Gold 6148 processor, 2.4GHz, 40 cores, turbo on, HT on, BIOS 86B.01.00.0412.R00, 12x16GB 2666MHz DDR, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. Gaussian* 16 (G16) AVX2 Certified Binaries. OS: Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. Testing by Intel April 2017. Workloads: apinehfb: Molecule: C10H16 , alpha-pinine, basic Hartree-Fock single-point, 678 basis functions, Route card: #P hf/6-311++G(3df,3pd) scf=tight apinefreqb: Molecule: C10H16 , alpha-pinine, frequency calculation single-point, 346 basis functions, Route card: #P rb3lyp/6-311G(df,p) freq hismp4: Molecule: C6H9N3O2 , histidine-H+ complex MP4, C1 symmetry, 183 basis functions, Route card: #P MP4/6-31G(d) SCF=noIncore BASELINE: 2S Intel® Xeon® processor CPU E5-2697 v4 , 2.3GHz, 36 cores, turbo on, HT off, BIOS 86B.01.01.0018, 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4 RDIMM, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. NEW: 2S Intel® Xeon® processor 6148 Gold, 2.4GHz, 40 cores, turbo on, HT off, BIOS 86B.01.00.0412, 192GB total memory, 12 slots / 16 GB / 2666 MT/s / DDR4 RDIMM, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327.. FIO POP*: POP workload. OS: CentOS 7.3.1611. Testing by Intel May 2017. BASELINE: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo and HT on, 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4, SATA3 4TB. NEW: 2S Intel® Xeon® Platinum processor 6148, 2.4GHz, 20 cores, turbo and HT on, 192GB total memory, 12 slots / 16GB / 2666 MT/s / DDR4, s3700 800G SSD. Intel Embree 2.16.1: BASELINE: 2 x Intel® Xeon® E5-2699 Processor v4 (Dual socket, 2.2GHz, 2 x 22 cores), 256 GB total memory, BIOS Configuration: default, turbo on, hyper-threading on, CentOS* release 7.2.1511 / 3.10.0- 327.36.2.el7.x86_64 NEW: 2 x Intel® Xeon® Platinum 8180 Processor (Dual socket, 2.5GHz, 2 x 28 cores), 192 GB total memory, BIOS Configuration: default, turbo on, hyper-threading on, Red Hat* Enterprise Linux Server 7.2 / 3.10.0-327.el7.x86_64 Itiviti Tbricks 2.10*: The critical path measured is from getting client FIX order to the first order sent out to the most attractive and liquid exchange. The test is performed in a controlled simulation of a highly volatile market. Simulated flow creates large bursts, several magnitudes higher than typical production load. Tested by Itiviti May 2017. BASELINE: 2S Intel® Xeon® processor E5-2699 v4, 2.2GHz, 22 cores, turbo on, HT off, 64GB total memory, @2400 MHz, Intel® SSD DC S3500 800GB, Oracle Linux* 6.7. NEW: Intel® Xeon® Platinum processor 8168, 2.7GHz, 24 cores, turbo on, HT off, 64GB total memory, @2666 MHz, Intel® SSD DC P3520 450GB, Oracle Linux* 6.7. Julia Computing: Ubuntu 16.04 (Xenial Xerus) kernel 4.4.0-75-generic Testing by Julia Computing May 2017. BASELINE: Intel® Xeon® processor E5-2699 v4, 2.2GHz, 44 cores, turbo and HT on, BIOS 338.R00, 128 GB total memory, (hard disk space), Ubuntu 16.04 (Xenial Xerus) kernel 4.4.0-75-generic NEW: Intel® Xeon® Platinum 8168 processor, 2.7 GHz, 48 cores, turbo and HT on, BIOS SE5C620.86B.01.00.0412.020920172159, 64 GB total memory, (Hard disk space), Ubuntu 16.04 (Xenial Xerus) kernel 4.4.0-75-generic
- 149. 149 FFT Actran* 1-2Mdof / unsymmetric complex – factorization workload only. OS: Red Hat Enterprise Linux* 7.3. Testing by Free Field Technologies April 2017. BASELINE: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 36 cores, turbo on, HT on, BIOS SE5C610.86B.01.01.0020.122820161512, 256GB total memory, 8 slots / 32GB / 2400 MT/s / DDR4 LRDIMM, 2x Seagate Constellation* ES.2 ST33000650NS, Red Hat Enterprise Linux* CentOS 7.3.1611 NEW: 2S Intel® Xeon® Gold 6148 processor, 2.4GHz, 40 cores, turbo on, HT on, BIOS SE5C620.86B.01.00.0412.020920172159, 192GB total memory, 12 slots / 16GB / 2666 MT/s / DDR4 LRDIMM, Intel® SSD DC S3700 + 3x Seagate* ST2000LX001 2TB, Red Hat Enterprise Linux* CentOS 7.3.1611 CONVERGE CFD 2.4.8 Custom Build for AVX2/AVX512. Tested by Intel, April 2017 BASELINE: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo and HT on, BIOS GRRFSDP1.86B0271.R00.1510301446, 128GB total memory, 8 memory channels / 8x16GB / 2400 MT/s / DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327.el7.x86_64, 2x Intel® SSD S3500 Series RAID0 array. NEW: 2S Intel® Xeon® Gold processor 6148, 2.4GHz, 20 cores, turbo boost tech on, turbo energy efficient off, HT on, BIOS SE5C620.86B.01.00.0412.020920172159, 192GB total memory, 12 memory channels / 12x16GB / 2666 MT/s / DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327.el7.x86_64, 2x Intel® SSD S3500 Series RAID0 array. HiFUN CONFIGURATION: HiFUN 4.1.1, Intel® Compiler 2016u3, Intel® MPI 5.1.3.181. - Compiler Flags: “-O3 –xCORE-AVX512 ” BASELINE: Dual Socket Intel® Xeon® processor E5-2697 v4, 2.3 GHz , 18 Cores/Socket, HT and turbo on, BIOS SE5C610.86B.01.01.0016.033120161139, 128 GB total memory, 8*16GB 2400 MHz DDR4, Red Hat Enterprise Linux 7.3, 800 GB SSD DC S3500 Series, Grantley-EP Broadwell-EP, Scalability tests performed on nodes with Intel® Omni-Path Architecture Si 100 Series, I_MPI_FABRICS=shm:tmi; I_MPI_TMI_PROVIDER=psm2 NEW: Dual socket Intel® Xeon® Gold 6148 processor, 2.4GHz, 20 cores/socket, EIST/Turbo on, SMT on, BIOS SE5C620.86B.01.00.0412.020920172159, 192 GB total memory, 12*16 GB DDR4, 1.2V RDIMM, 800GB IntelSSDSC2BA80, Red Hat Enterprise Linux 7.3, Wolf Pass, kernel 3.10.0-514.6.2.0.1.el7.x86_64.knl1, Scalability tests performed on nodes with Intel® Omni-Path Architecture Si 100 Series, I_MPI_FABRICS=shm:tmi; I_MPI_TMI_PROVIDER=psm2. KAIST/KISTI DNS-TBL CONFIGURATION: Intel® Fortran Compiler 17.0 Update 4, Intel® Math Kernel Library (Intel® MKL) 2017 Update 3, and Intel® MPI Library 2017 Update 3 - Compiler Flags: “-O3 –xCORE-AVX512 -align array64byte -r8 -qopenmp -assume buffered_io -mkl=parallel” BASELINE: Dual Socket Intel® Xeon® processor E5-2697 v4, 2.3 GHz , 18 Cores/Socket, HT on, turbo off, BIOS SE5C610.86B.01.01.0016.033120161139, 128 GB total memory, 8*16GB 2400 MHz DDR4, Red Hat* Enterprise Linux 7.3, Intel® OPA driver 10.2.0.0.158_72, tmi (psm2) fabric NEW: Dual socket Intel® Xeon® Gold 6148 processor, 2.4GHz, 20 cores/socket, HT on, Turbo off, BIOS SE5C620.86B.01.00.0412.020920172159, 192 GB total memory, 12*16 GB DDR4, 1.2V RDIMM, Red Hat Enterprise Linux 7.3 , Intel® OPA driver 10.2.0.0.158_72, tmi (psm2) fabric Xidian University: MOM-LASPCEM*: OS: CentOS* 7.3.1611. Testing by Intel May 2017. BASELINE : Intel® Xeon® Gold 6148 processor, 2.4GHz, 20 cores, turbo and HT on, 192GB total memory, 12 slots / 16GB / 2666 MT/s / DDR4, Intel® SSD DC P3700 1.6TB. NEW (for Intel® Optane™ SSD benchmarking): Intel® Xeon® Gold processor 6148, 2.4GHz, 20 cores, turbo and HT on, 192GB total memory, 12 slots / 16GB / 2666 MT/s, Intel® Optane™ SSD DC P4800X, 375GB. LAMMPS: BASELINE: 2S Intel® Xeon® processor E7-2697 v3, 2.6GHz, 28 cores, Intel® Turbo Boost Technology and Intel® Hyper-Threading technology on, BIOS 86B.01.01.1008.R00, 8x8GB 2133 MHz DDR4, CentOS Linux* 7.1.1503 kernel 3.10.0-229. NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 36 cores, Intel® Turbo Boost Technology and Intel® Hyper-Threading technology on, BIOS 86B0271.R00, 8x16GB 2400MHz DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. NEW: 2S Intel® Xeon® Gold 6148 processor, 2.4GHz, 40 cores, Intel® Turbo Boost Technology and Intel® Hyper-Threading technology on, BIOS 86B.01.00.0412.R00, 12x16GB 2666MHz DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. GROMACS AVX2 CONFIGURATION: Version 2016.3: ftp://ftp.gromacs.org/pub/gromacs/gromacs-2016.3.tar.gz , Intel® Compiler 17.0.1.132, Intel® MPI 2017u1. Optimization Flags: “-O3 -xCORE-AVX2“. Cmake options: “-DGMX_FFT_LIBRARY=mkl -DGMX_SIMD=AVX2_256”. GROMACS AVX512 CONFIGURATION: Version 2016.3: ftp://ftp.gromacs.org/pub/gromacs/gromacs-2016.3.tar.gz , Intel® Compiler 17.0.1.132, Intel® MPI 2017u1. Optimization Flags: “-O3 -xCORE-AVX512“. Cmake options: “-DGMX_FFT_LIBRARY=mkl -DGMX_SIMD=AVX_512”. BASELINE CONFIGURATION: GROMACS AVX2 binary, Dual Socket Intel® Xeon® processor E5-2697 v3 2.6 GHz, 14 Cores/Socket, 28 Cores, 56 Threads (HT on, Turbo on), DDR4 128GB, 2133 MHz, Red Hat* 7.3. NEXT GEN CONFIGURATION: GROMACS AVX2 binary, Dual Socket Intel® Xeon® processor E5-2697 v4 2.3 GHz, 18 Cores/Socket, 36 Cores, 72 Threads (HT on, Turbo on), DDR4 128GB, 2400 MHz, Red Hat 7.2. NEW CONFIGURATION: GROMACS AVX512 binary, Dual Socket Intel® Xeon® Gold 6148 processor 2.4 GHz , 20 Cores/Socket, 40 Cores, 80 Threads (HT on, Turbo on), DDR4 192GB, 2666 MT/s DDR4 RDIMMs, Red Hat 7.2.
- 150. 150 VASP CONFIGURATION: BetaVASP, a release candidate for v6.0. Developer branch provided as “Package” included with download: https://github.com/vasp-dev/vasp-knl. AVX512: Intel® Compiler 17.0.1.132, Intel® MPI 2017u1, ELPA 2016.05.004. Optimization Flags: “-O3 -xCORE-AVX512“. AVX2: Intel® Compiler 17.0.1.132, Intel® MPI 2017u1, ELPA 2016.05.004. Optimization Flags: “-O3 -xCORE-AVX2“. BASELINE CONFIGURATION: 2S Intel® Xeon® processor E5-2699 v3 2.3 GHz, 18 Cores/Socket, 36 Cores, 72 Threads, HT on, turbo off, 128GB total memory, 2133 MT/s / DDR4 RDIMM, Red Hat Enterprise Linux* 7.0 kernel. NEXT GEN CONFIGURATION: 2S Intel® Xeon® processor E5-2697 v4 2.3 GHz , 18 Cores/Socket, 36 Cores, 72 Threads, HT on, turbo off, BIOS 86B0271.R00, 128GB total memory, 2400 MT/s DDR4 RDIMM, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. NEW CONFIGURATION: 2S Intel® Xeon® Gold 6148 processor 2.4 GHz , 20 Cores/Socket, 40 Cores, 80 Threads, HT on, turbo off, BIOS 86B.01.00.0412, 192GB total memory, 2666 MT/s / DDR4 RDIMM, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. Amber: Version 16 with all patches applied at December, 2016. Workloads: PME Cellulose NVE(408K atoms), PME stmv(1M atoms), GB Nucleosome (25K), GB Rubisco (75K). No cut-off used for GB workloads. Compiled with -mic2_spdp –intelmpi - openmp, –DMIC2 defined. DMIC2 enable optimization for AVX512 vectorization, SPDP mixed precision, OpenMP optimization.Tests performed on March 2017. BASELINE: Executed with 36 MPI, 2 OpenMP. 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 36 cores, turbo and HT on, BIOS 86B0271.R00, 8x16GB 2400MHz DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. NEW: Executed with 40 MPI and 2 OpenMP. 2S Intel® Xeon® Gold 6148 processor, 2.4GHz, 40 cores, turbo on, HT on, BIOS 86B.01.00.0412.R00, 12x16GB 2666MHz DDR, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. NAMD: Version 2.12 Dec2016. Workloads: apoa1(92K atoms), stmv(1M atoms). Compiled with –DNAMD_KNL* define. Tests performed on March 2017. BASELINE: Executed with 72 charm threads. 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 36 cores, turbo and HT on, BIOS 86B0271.R00, 8x16GB 2400MHz DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. Compiler option “-xCORE-AVX2”. NEW: Executed with 40 charm threads. 2S Intel® Xeon® Gold 6148 processor, 2.4GHz, 40 cores, turbo on, HT on, BIOS 86B.01.00.0412.R00, 12x16GB 2666MHz DDR, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. Compiler option “-xCORE-AVX512”. QMCPACK CONFIGURATION: Developer branch provided as “rebuild-dirac.soa-fix-distAA” downloaded from: https://xgitlab.cels.anl.gov/QMCPACK/qmcpack-dev.git, Date: 04/05/2017. Intel® parallel studio 17.2.174, Optimization Flags: “-o3 -ip -restrict -unroll -g -debug inline-debug-info -openmp -std=c++11”. BASELINE CONFIGURATION: 2S Intel® Xeon® processor E5-2697 v4 2.3 GHz , 18 Cores/Socket, 36 Cores, 72 Threads, HT on, turbo off, BIOS 86B0271.R00, 128GB total memory, 2400 MT/s DDR4 RDIMM, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. (-xCORE-AVX2), run with 2088 walkers, 2 MPI with 36 threads each. NEW CONFIGURATION: 2S Intel® Xeon® processor Gold 6148 processor 2.4 GHz , 20 Cores/Socket, 40 Cores, 80 Threads, HT on, turbo on, BIOS 86B.01.00.0412, 192GB total memory, 2666 MT/s DDR4 RDIMM, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. (-xCORE-AVX512), run with 2080 walkers, 2 MPI with 40 threads each. Portable code transformation to increase SIMD efficiency and reduce memory footprint [1,2] The efforts result in 3.1x-4.3x improvements on 2S Intel® Xeon® processor Gold 6148 and 2.6x-6.3x improvements on 2S Intel® Xeon® processor E5-2697 v4 processor from baseline code. These speedups are wrt. QMPCACK release 3.0.0 on the same hardware with QMC_MIXED_PRECISION=0 flag. Monte Carlo workload. OS: Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. Testing by Intel March 2017. BASELINE: 2S Intel® Xeon® processor E5-2697 v3, 2.6GHz, 28 cores, turbo and HT on, BIOS 86B.0036.R05, 64GB total memory, 8x8GB 2133 MHz DDR4, Fedora release 20 kernel 3.15.10-200 . NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4 , 2.3GHz, 36 cores, turbo and HT on, BIOS 86B0271.R00, 128GB total memory, 8 x16GB 2400 MHz DDR4 RDIMM, 1 x 1TB SATA, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. NEW: 2S Intel® Xeon® Gold 6148 processor @ 2.4GHz, H0QS, 40 cores 150W. QMS1, turbo and HT on, BIOS SE5C620.86B.01.00.0412.020920172159, 192GB total memory, 12 x 16 GB 2666 MHz DDR4 RDIMM, 1 x 800GB Intel® SSD SC2BA80, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327 Black-Scholes workload. OS: Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. Testing by Intel March 2017. BASELINE: 2S Intel® Xeon® processor E5-2697 v3, 2.6GHz, 28 cores, turbo and HT on, BIOS 86B.0036.R05, 64GB total memory, 8x8GB 2133 MHz DDR4, Fedora release 20 kernel 3.15.10-200 . NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4 , 2.3GHz, 36 cores, turbo and HT on, BIOS 86B0271.R00, 128GB total memory, 8 x16GB 2400 MHz DDR4 RDIMM, 1 x 1TB SATA, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. NEW: 2S Intel® Xeon® Gold 6148 processor @ 2.4GHz, H0QS, 40 cores 150W. QMS1, turbo and HT on, BIOS SE5C620.86B.01.00.0412.020920172159, 192GB total memory, 12 x 16 GB 2666 MHz DDR4 RDIMM, 1 x 800GB Intel® SSD SC2BA80, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327
- 151. 151 Binomial workload. OS:Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. Testing by Intel March 2017. BASELINE: 2S Intel® Xeon® processor E5-2697 v3, 2.6GHz, 28 cores, turbo and HT on, BIOS 86B.0036.R05, 64GB total memory, 8x8GB 2133 MHz DDR4, Fedora release 20 kernel 3.15.10-200 . NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4 , 2.3GHz, 36 cores, turbo and HT on, BIOS 86B0271.R00, 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4 RDIMM, 1 x 1TB SATA, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. NEW: 2S Intel® Xeon® Gold processor 6148, 2.4GHz, 40 cores, turbo and HT on, BIOS 86B.01.00.0412, 192GB total memory, 12 slots / 16 GB / 2666 MT/s / DDR4 RDIMM, 1 x 800GB Intel® SSD SC2BA80, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. Berkeley GW 1.2 Sigma phase MTXEL-COR calculation of Benzene. OS: Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. Testing by Intel March 2017. BASELINE: 2S Intel® Xeon® processor E5-2697 v4 , 2.3GHz, 36 cores, turbo and HT on, BIOS 86B0271.R00, 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4 RDIMM, 1 x 1TB SATA, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. NEW: 2S Intel® Xeon® Gold 6148 processor, 2.4GHz, 40 cores, turbo and HT on, BIOS 86B.01.00.0412, 192GB total memory, 12 slots / 16 GB / 2666 MT/s / DDR4 RDIMM, 1 x 800GB Intel® SSD SC2BA80, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. PPPL GTC-P: BASELINE: 2S Intel® Xeon® processor E5-2699 v3, 2.3GHz, 18 cores, turbo and HT off, Red Hat Enterprise Linux* 7.2-Linux-3.10.0-123 & 2S Intel® Xeon® processor E3-2697 v4, 2.3GHz, 18 cores, turbo and HT on, Red Hat Enterprise Linux* 7.2-Linux-3.10.0-514.6.2.0.1 NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo and HT on, Red Hat Enterprise Linux* 7.2-Linux-3.10.0-514.6.2.0.1 NEW: 2S Intel® Xeon® Gold 6148 processor, 2.4GHz, 20 cores, turbo and HT on, Red Hat Enterprise Linux* 7.2-Linux-3.10.0-514 MILC version: https://github.com/milc-qcd/milc_qcd (master branch). Workload: ks_imp_rhmc. The “ks_imp_rhmc” is a dynamical RHMC (rational hybrid Monte Carlo algorithm) code for staggered fermions Workload/Binary: ‘ks_imp_rhmc’ compiled with Makefile included with MILC package (modified to enable QPhiX and AVX512). Used Intel® Compiler and Intel® MPI part of Intel® Parallel Studio XE Cluster Edition 2017 update 2. Run Configuration: Lattice Volume: 24^(4). Runs performed with optimal MPI x OpenMP configuration for the respective processors. KMP_AFFINITY=granularity= fine, scatter. BASELINE: 2S Intel® Xeon® processor E5-2699 v3, 2.3GHz, 36 cores, turbo and HT on, BIOS 86B.0036.R05, 128GB total memory, 8 slots / 16GB / 2133 MT/s / DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-123 NEXT GEN: 2S Intel® Xeon® processor CPU E5-2697 v4 , 2.3GHz, 36 cores, turbo and HT on, BIOS 86B0271.R00, 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4 RDIMM, 1 x 1TB SATA, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. NEW: 2S Intel® Xeon® Gold processor 6148, 2.4GHz, 40 cores, turbo and HT on, BIOS 86B.01.00.0412, 192GB total memory, 12 slots / 16 GB / 2666 MT/s / DDR4 RDIMM, 1 x 800GB Intel® SSD SC2BA80, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. SpecMPI2007 Configuration: Intel Compiler 17.1.132, Intel MPI 17.1.132; Optimization FLAGS: “-O3 -xCORE-AVX512 -no-prec-div”. Run flags: I_MPI_COMPATIBILITY=3, I_MPI_FABRICS=shm:tmi. BASELINE: 2S Intel® Xeon® processor E7-2697 v3, 2.6GHz, 28 cores, turbo and HT on, BIOS 86B.01.01.1008.R00, 8x8GB 2133 MHz DDR4, CentOS Linux* 7.1.1503 kernel 3.10.0-229. NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 36 cores, turbo and HT on, BIOS 86B0271.R00, 8x16GB 2400MHz DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. NEW: 2S Intel® Xeon® Gold 6148 processor, 2.4GHz, 40 cores, turbo and HT on, BIOS 86B.01.00.0412.R00, 12x16GB 2666MHz DDR, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. HOMME version: https://svn-homme-model.cgd.ucar.edu/branch_tags/dungeon_tags/dungeon016. Compiled with “-O3 -fp-model fast -xCORE-AVX2”. Running “perfTestWACCM” benchmark from code repository with size NE=8. Executed with 64 MPI ranks and OMP_NUM_THREADS=1. BASELINE: 2S Intel® Xeon® processor CPU E5-2697 v4 , 2.3GHz, 36 cores, turbo and HT on, BIOS 86B0271.R00, 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4 RDIMM, 1 x 1TB SATA, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. NEW: 2S Intel® Xeon® Gold processor 6148, 2.4GHz, 40 cores, turbo and HT on, BIOS 86B.01.00.0412, 192GB total memory, 12 slots / 16 GB / 2666 MT/s / DDR4 RDIMM, 1 x 800GB Intel® SSD SC2BA80, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. OMP2012: Software: Compiled using Intel compiler “-O3 -xCORE-AVX512 –qopenmp –ansi-alias”. Executed with OMP_NUM_THREADS=80 for base and 40/80 for peak. BASELINE: 2S Intel® Xeon® processor CPU E5-2697 v4 , 2.3GHz, 36 cores, turbo and HT on, BIOS 86B0271.R00, 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4 RDIMM, 1 x 1TB SATA, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. Software: Compiled using Intel compiler “-O3 -xCORE-AVX2 –qopenmp –ansi-alias”. Executed with OMP_NUM_THREADS=72 for base and 36/72 for peak. NEW: 2S Intel® Xeon® Gold processor 6148, 2.4GHz, 40 cores, turbo and HT on, BIOS 86B.01.00.0412, 192GB total memory, 12 slots / 16 GB / 2666 MT/s / DDR4 RDIMM, 1 x 800GB Intel® SSD SC2BA80, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327.
- 152. 152 CP2K v5.0 (Development),git clone --branch intel https://github.com/cp2k/cp2k.git cp2k.git (603bcd0). Workloads: H2O-dft-ls (set NREP=3 in tests/QS/benchmark_DM_LS/H2O-dft-ls.inp), libtest (modified tests/LIBTEST/dbcsr_blocks_04.inp), and H2O-64 (tests/QS/benchmark/H2O-64.inp). Binary: see https://github.com/hfp/xconfigure/tree/master/cp2k. Intel® Parallel Studio XE Cluster Edition for Linux (Update 4), LIBXSMM 1.8.1, LIBINT 1.1.6, and LIBXC 3.0.0, and ELPA 2017.05.001.rc1. OS: Red Hat Enterprise Linux* 7.3 kernel 3.10. Testing by Intel in June 2017 using the same version of CP2K, the same build settings, and development tool. Run Configuration: mpirun -bootstrap ssh -genvall -np $(NRANKS) -genv I_MPI_PIN_DOMAIN=auto -genv KMP_AFFINITY=scatter,granularity=fine,1 -genv OMP_NUM_THREADS=${NTHREADS_PER_RANK} exe/Linux-x86-64-intel/cp2k.psmp workload.inp BASELINE: 2S Intel® Xeon® E5-2699 v3, 2.3 GHz, 2x18 cores, turbo+HT+CoD on, 64 GB mem. (2133 MT/s) NEXT GEN: 2S Intel® Xeon® E5-2697 v4, 2.3 GHz, 2x18 cores, turbo+HT on, 128 GB total mem. (2400 MT/s) NEW: 2S Intel® Xeon® Gold 6148 processor, 2.4 GHz, 2x20 cores, turbo+HT on, 192 GB total memory (2666 MT/s) specfem3D_globe benchmarks tested by Intel, June 2017, 1 node measurement BASELINE: 2S Intel® Xeon® processor E5-2697 v3, 2.6GHz, 14 cores, turbo off and HT on, 128GB total memory, 8 memory channels / 8x16GB / 2133 MT/s / DDR4, Red Hat Enterprise Linux* 7.3, kernel 3.10.0- 514.6.2.0.1.el7.x86_64.knl1 NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo off and HT on, BIOS GRRFSDP1.86B0271.R00.1510301446, 128GB total memory, 8 memory channels / 8x16GB / 2400 MT/s / DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-514.6.2.0.1.el7.x86_64.knl1. NEW: 2S Intel® Xeon® Gold 6148 processor, 2.4GHz, 20 cores, turbo boost tech off, turbo energy efficient off, HT on, BIOS SE5C620.86B.01.00.0412.020920172159, 192GB total memory, 12 memory channels / 12x16GB / 2666 MT/s / DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-514.6.2.0.1.el7.x86_64.knl1. QPhiX benchmarks tested by Intel, May 2017, 1 node measurement BASELINE: 2S Intel® Xeon® processor E5-2697 v3, 2.6GHz, 14 cores, turbo and HT on, 128GB total memory, 8 memory channels / 8x16GB / 2133 MT/s / DDR4, Red Hat Enterprise Linux* 7.3, kernel 3.10.0- 514.6.2.0.1.el7.x86_64.knl1 NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo and HT on, BIOS GRRFSDP1.86B0271.R00.1510301446, 128GB total memory, 8 memory channels / 8x16GB / 2400 MT/s / DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-514.6.2.0.1.el7.x86_64.knl1. NEW: 2S Intel® Xeon® Gold 6148 processor, 2.4GHz, 20 cores, turbo boost tech on, turbo energy efficient off, HT on, BIOS SE5C620.86B.01.00.0412.020920172159, 192GB total memory, 12 memory channels / 12x16GB / 2666 MT/s / DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-514.6.2.0.1.el7.x86_64.knl1. Chroma benchmarks tested by Intel, May 2017, 1 node measurement BASELINE: 2S Intel® Xeon® processor E5-2697 v3, 2.6GHz, 14 cores, turbo and HT on, 128GB total memory, 8 memory channels / 8x16GB / 2133 MT/s / DDR4, Red Hat Enterprise Linux* 7.3, kernel 3.10.0- 514.6.2.0.1.el7.x86_64.knl1 NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo and HT on, BIOS GRRFSDP1.86B0271.R00.1510301446, 128GB total memory, 8 memory channels / 8x16GB / 2400 MT/s / DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-514.6.2.0.1.el7.x86_64.knl1. NEW: 2S Intel® Xeon® Gold 6148 processor, 2.4GHz, 20 cores, turbo boost tech on, turbo energy efficient off, HT on, BIOS SE5C620.86B.01.00.0412.020920172159, 192GB total memory, 12 memory channels / 12x16GB / 2666 MT/s / DDR4, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-514.6.2.0.1.el7.x86_64.knl1. ANL HACC*: 2563 N-body particles. OS: Red Hat Enterprise Linux* 7.2 kernel 3.10. Testing by Intel March 2017. BASELINE: 2S Intel® Xeon® processor CPU E5-2697 v4 , 2.3GHz, 36 cores, turbo and HT on, BIOS 86B0271.R00, 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4 RDIMM, 1 x 1TB SATA, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. NEW: 2S Intel® Xeon® Gold processor 6148, 2.4GHz, 40 cores, turbo and HT on, BIOS 86B.01.00.0412, 192GB total memory, 12 slots / 16 GB / 2666 MT/s / DDR4 RDIMM, 1 x 800GB Intel® SSD SC2BA80, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. Mines ParisTech Fwi2d modeling & Fwi2d inversion: version dev1.6 from june16. OS: Centos 7.3. Compiler: Intel® Parallel Studio XE Cluster Edition 2017 update 2. Run config: MPI+OMP . One MPI per socket. Same workload for all IA. Workload: Extended Marmousi-1 , single-node problem-size: 600x1423. Testing by Intel, April 2017 BASELINE: 2S Intel® Xeon® processor E5-2680 v3, 2.5GHz, 24 cores (12 each socket), turbo and HT OFF. DDR4 2133 Mhz NEXT GEN: 2S Intel® Xeon® processor E5-2680 v4, 2.4GHz, 28 cores (14 each socket), turbo and HT OFF. DDR4 2400 Mhz NEW: 2S Intel® Xeon® Gold 6148 processor, 2.4GHz, 40 cores (20 each socket), turbo and HT OFF. DDR4 2667 Mhz
- 153. 153 iso3dfd*: version dev13from Jan 2017. OS: Centos 7.3. Compiler: Intel® Parallel Studio XE Cluster Edition 2017 update 2. Runs: OpenMP only using always the max number of cores. Common workload parameters for ALL runs. Workload: Homogeneous model , single-node problem-size: 224x212x2100. Testing by Intel, April 2017 BASELINE: 2S Intel® Xeon® processor E5-2680 v3, 2.5GHz, 24 cores (12 each socket), turbo and HT OFF. DDR4 2133 Mhz NEXT GEN: 2S Intel® Xeon® processor E5-2680 v4, 2.4GHz, 28 cores (14 each socket), turbo and HT OFF. DDR4 2400 Mhz NEW: 2S Intel® Xeon® Gold 6148 processor, 2.4GHz, 40 cores (20 each socket), turbo and HT OFF. DDR4 2667 Mhz YASK software release v20170421 AWP, single-node problem-size: 1024*1024*128 (approx. 13.6 GB). Testing by Intel, April 2017 BASELINE: 2S Intel® Xeon® processor E5-2699 v3, 2.6GHz, 36 cores (18 each socket), turbo and HT on, 128 GB total memory, Red Hat* Enterprise Linux Server release 7.0 (Maipo), 3.10.0-123.el7.x86_64 kernel NEXT GEN: 2S Intel® Xeon® processor E5-2697 v4, 2.3GHz, 36 cores (18 each socket), turbo and HT on, 256 GB total memory, Red Hat Enterprise Linux Server release 7.2 (Maipo), 3.10.0-327.el7.x86_64 kernel NEW: 2S Intel® Xeon® Gold 6148 processor, 2.4GHz, 40 cores (20 each socket), turbo and HT on, 192 GB total memory, Red Hat Enterprise Linux Server release 7.2 (Maipo), 3.10.0-327.el7.x86_64 kernel WRF: BASELINE: 2S Intel® Xeon® processor CPU E5-2697 v4 , 2.3GHz, 36 cores, turbo and HT on, BIOS 86B0271.R00, 128GB total memory, 8 slots / 16GB / 2400 MT/s / DDR4 RDIMM, 1 x 1TB SATA, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. Software: WRF version 3.6.1 Compiled using Intel config option with “-O3 -fp-model fast=1 -xCORE-AVX2”. Executed with 36 MPI ranks and OMP_NUM_THREADS=1. NEW: 2S Intel® Xeon® Gold processor 6148, 2.4GHz, 40 cores, turbo and HT on, BIOS 86B.01.00.0412, 192GB total memory, 12 slots / 16 GB / 2666 MT/s / DDR4 RDIMM, 1 x 800GB Intel® SSD SC2BA80, Red Hat Enterprise Linux* 7.2 kernel 3.10.0-327. Software: WRF version 3.6.1 Compiled using Intel config option with “-O3 -fp-model fast=1 -xCORE-AVX512”. Executed with 40 MPI ranks and OMP_NUM_THREADS=1. MPAS-O CONFIGURATION: 27 Jul 2016 (Git Hash: v4.0-1313-g615ede1), Intel® Compiler 2016u4, Intel® MPI 5.1.3.181. Building all the 3rd party libraries (NetCDF, ParallelNetCDF, PIO) with “ –O3 –xCORE- AVX512” while compiling on SKX. Optimization Flags: “-O3 –xCORE-AVX512”, run times exclude time to I/O, I/O frequency set to > 2 days to exclude I/O dumps. Command: make ifort CORE=ocean BASELINE: Dual Socket Intel® Xeon® processor E5-2697 v4, 2.3 GHz , 18 Cores/Socket, HT and turbo on, BIOS SE5C610.86B.01.01.0016.033120161139, 128 GB total memory, 8*16GB 2400 MHz DDR4, Red Hat 7.2, 800 GB SSD DC S3500 Series, Grantley-EP Broadwell-EP, Scalability tests performed on nodes with Intel® Omni-Path Si 100 Series, I_MPI_FABRICS=shm:tmi; I_MPI_TMI_PROVIDER=psm2 NEW: Dual socket Intel® Xeon® Gold 6148 processor, 2.4GHz, 20 cores/socket, EIST/Turbo on, SMT on, BIOS SE5C620.86B.01.00.0412.020920172159, 192 GB total memory, 12*16 GB DDR4, 1.2V RDIMM, 800GB IntelSSDSC2BA80, Red Hat Enterprise Linux* 4.8.5-11, Wolf Pass, kernel 3.10.0-514.6.2.0.1.el7.x86_64.knl1, Scalability tests performed on nodes with Intel® Omni-Path Si 100 Series, I_MPI_FABRICS=shm:tmi; I_MPI_TMI_PROVIDER=psm2. IBM DB2 2x2: Throughput of 3TB BDInsights analytics workload (Queries per Hour). Testing by Intel, April 2017 BASELINE: IBM DB2 v10.5.0.8, OS: SLES 11.3, 2S Intel® Xeon® processor E7-4890 v2, 1.5TB (96 DIMMs / 16GB / 1066MHz / DDR3 DIMM), IBM® XIV storage (132 10K disks) NEW: IBM DB2 v11.1.1.1, OS: RHEL 7.3, 2S Intel® Xeon® Platinum 8180 Processor, 1.5TB (24 DIMMs / 16GB / 2666 MHz / DDR4 DIMM), 1x Intel® SSD DC S3700 Aerospike Database 2x2: The database was populated with 200 M records of 100 bytes each and benchmarked with the Aerospike Java Benchmark tool (https://github.com/aerospike/aerospike-client-java) . The workload simulated 95%/5% read/update ratio. Two Aerospike instances were launched on a single server forming a cluster. Each Aerospike instance was affinitized to a CPU socket and configured to use one of the 10GB NICs. Each 10GB NIC had its interrupt IRQs affinitized to a CPU socket. BASELINE: Aerospike Server Enterprise* 3.6.4 , CentOS* 6.7, kernel version 2.6.32-573.3.1.el6.x86_64, 2 Intel® Xeon® processor E5-2697 v3, 2.6GHz, 28 cores, 128GB DDR4/1866, regular DIMM, 2x 10Gb network Intel X540-AT2 not bonded, no disk used – in memory workload, Clients: 8 client systems were used to concurrently submit queries to the servers and drive the workload. The same clients were used in both “baseline” and “new”. The clients were configured as follows: : E5- 2697 v3 128GB of memory and 10GB Intel X540-AT2 network The database was populated with 400 M records of 100 bytes each and benchmarked with the Aerospike Java Benchmark tool (https://github.com/aerospike/aerospike-client-java) . The workload simulated 95%/5% read/update ratio. Two Aerospike instances were launched on a single server forming a cluster. NEW: Aerospike Server Enterprise 3.12.1, OS: CentOS 7.2 with kernel updated to 4.4.59, Intel® Xeon® Platinum processor 8180, 2.5GHz, 28 cores, turbo and HT on, BIOS SE5C620.86B.01.00.0412.020920172159 , 384GB total memory, 12 DIMMs / 32GB / Configured Clock Speed: 2666 MHz / DDR4 DIMM, 2 x Intel® 82599ES 10 Gigabit Ethernet Controllers – all 4 ports on the 2 network controllers were bonded for an aggregate 40000Mb/s bond. No storage – in-memory workload Clients: 8 client systems were used to concurrently submit queries to the servers and drive the workload. The same clients were used in both “baseline” and “new”. The clients were configured as follows: CentOS 7.2 with kernel 3.10.0-327. Intel® Xeon® processor E5-2697 v4, 2.3GHz, 18 cores, turbo and HT on, BIOS SE5C610.86B.01.01.0016.033120161139, 128GB total memory, 8 DIMMs / 16GB / Configured Clock Speed: 2400 MHz, 1 x Intel® 82599ES 10 Gigabit Ethernet Controllers
- 154. 154 LegalNoticesandDisclaimers Intel technologies’ featuresand benefits depend on system configuration and may require enabled hardware, software or service activation. Learn more at intel.com, or from the OEM or retailer. No computer system can be absolutely secure. Tests document performance of components on a particular test, in specific systems. Differences in hardware, software, or configuration will affect actual performance. Consult other sources of information to evaluate performance as you consider your purchase. For more complete information about performance and benchmark results, visit http://www.intel.com/performance. Normalized performance is calculated by assigning a baseline value of 1.0 to one benchmark result, and then dividing the actual benchmark result for the baseline platform into each of the specific benchmark results of each of the other platforms, and assigning them a relative performance number that correlates with the performance improvements reported. Cost reduction scenarios described are intended as examples of how a given Intel-based product, in the specified circumstances and configurations, may affect future costs and provide cost savings. Circumstances will vary. Intel does not guarantee any costs or cost reduction. This document contains information on products, services and/or processes in development. All information provided here is subject to change without notice. Contact your Intel representative to obtain the latest forecast, schedule, specifications and roadmaps. No license (express or implied, by estoppel or otherwise) to any intellectual property rights is granted by this document. Intel does not control or audit third-party benchmark data or the web sites referenced in this document. You should visit the referenced web site and confirm whether referenced data are accurate. Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. Intel, Xeon, AVX-512, Omni-Path Architecture, Quick Assist, Optane, Trusted Execution Technology, Key Protection Technology, Platform Trust Technology, Resource Director Technology, the Intel logo and others are trademarks of Intel Corporation and its subsidiaries in the U.S. and/or other countries. *Other names and brands may be claimed as the property of others. © 2017 Intel Corporation.
- 155. Copyright © 2017,Intel Corporation. All rights reserved.