SlideShare a Scribd company logo

PCCC23:日本AMD株式会社 テーマ2「AMD EPYC™ プロセッサーを用いたAIソリューション」

PCクラスタコンソーシアムの開催するPCCC23(第23回PCクラスタシンポジウム)「HPC基盤技術と生成AI」会員展示として公開中のスライドです。 詳細は下記のWebサイトへお越しください。 ▽PCCC23 「HPC基盤技術と生成AI」 (2023年12月7日~8日) https://www.pccluster.org/ja/event/pccc23/

1 of 12
Download to read offline
AMD EPYC™
Processors for AI
Nov 2023
2 |
[Public]
Broad Industry Impact
Self-driving cars can
recognize signage,
pedestrians, and
other vehicles to be
avoided
Monitor quality of
manufactured
products from food
items to printed
circuit boards
Automate checkout
lines and use product
recommendation
engines to offer
suggestions, whether
online or in
the store
Detect anomalies
including fractures
and tumors.
Use the same models
in research to assess
in vitro cell growth
and proliferation
Natural language
processing can use
spoken requests and
recommendation
engines to help point
customers to
solutions
AI-powered anomaly
detection helps stop
credit-card fraud,
while computer vision
models watch for
suspicious documents
Automotive Manufacturing Retail
Financial
Services
Medical
Service
Automation
AI extends and enriches common business workloads and activities
3 |
[Public]
AMD Propels theAI Lifecycle
TRAINING
The most data- and processing-intensive
part of the AI lifecycle. Significant
computing power is required, and servers
equipped with AMD Instinct™ accelerators
are designed to accelerate the process
INFERENCING
Once trained, AI requires comparatively
less processing power to process
incoming data and business records in
real time. Inferencing happens close to
the data and AMD EPYC™ processors
are ideal for inferencing.
4 |
[Public]
AMD EPYC™ Processors:
Inference Performance
1.78x
Up
to
SERVERS BASED ON AMD EPYC™ 9654 CPUS
RECOGNIZE VEHICLES AT 1.78X THE RATE OF
INTEL® XEON® Platinum 8940H CPU BASED SERVERS
Phoronix used the OpenVINO benchmark using INT8-FP16 data types to compare multiple CPU types. They measured a whopping 78% speedup on vehicle detection,
and a 14% speedup on age-gender recognition comparing a 2P Intel Xeon 8490H processor-powered server to a 2P AMD EPYC 9654 processor-powered server with
ATX-512 on. See endnotes SP5-192, -193
6207
11029
0 2000 4000 6000 8000 10000 12000
2 x Intel Xeon Platinum 8940H
2 x AMD EPYC 9654
OpenVINO FP16-INT8
Vehicle Detection FPS
(Higher is Better)
103184
118104
0 20000 40000 60000 80000 100000 120000 140000
2 x Intel Xeon Platinum 8940H
2 x AMD EPYC 9654
OpenVINO FP16-INT8
Age Gender Recognition Faces per Second
(Higher is Better)
5 |
[Public]
256 Threads for End-to-EndAI Boost
Results may vary due to factors including system configurations, software versions and BIOS settings. As of 6/13/2023, see endnotes: SP5-051.
• Comparison derived from TPCx-AI benchmark
covering 10 end-to-end use cases covering
training, serving and throughput
• 128C AMD EPYC 9754 delivers up to an aggregate
of ~2.2x the AI test cases per min. vs. 60C Intel
Xeon Platinum 8490H
Outstanding end-to-end AI throughput
performance on a wide variety of use cases
831
1841
120 total cores/
240 threads
4 instances / 30 vCPUs per
256 total cores/
512 threads
8 instances / 30 vCPU per
Xeon® Platinum 8490H AMD EPYC™ 9754
~2.2x
Running 2Pserverswith 128C4thGenAMDEPYC™ 9754vs.60C4thGenIntel® Xeon® Platinum8490H
End-to-end AI data science pipeline
aggregate AI
use cases/min
6 |
[Public]
ROCm™ Platform Vitis™ AI Platform
CPU Stack
Unified Inferencing Model StreamlinesAdoption
The Unified Inference Frontend (blue) provides a uniform way to link your inferencing software with the acceleration capabilities of
EPYC™ CPUs, AMD Instinct™ accelerators, and Versal™ and Zynq™ adaptive SoCs
The CPU-specific software stack includes a robust set of tools that accelerate deep learning and inference workloads

Recommended

Intel and DataStax: 3D XPoint and NVME Technology Cassandra Storage Comparison
Intel and DataStax: 3D XPoint and NVME Technology Cassandra Storage ComparisonIntel and DataStax: 3D XPoint and NVME Technology Cassandra Storage Comparison
Intel and DataStax: 3D XPoint and NVME Technology Cassandra Storage ComparisonDataStax Academy
 
Yashi dealer meeting settembre 2016 tecnologie xeon intel italia
Yashi dealer meeting settembre 2016 tecnologie xeon intel italiaYashi dealer meeting settembre 2016 tecnologie xeon intel italia
Yashi dealer meeting settembre 2016 tecnologie xeon intel italiaYashi Italia
 
Building Efficient Edge Nodes for Content Delivery Networks
Building Efficient Edge Nodes for Content Delivery NetworksBuilding Efficient Edge Nodes for Content Delivery Networks
Building Efficient Edge Nodes for Content Delivery NetworksRebekah Rodriguez
 
Performance out of the box developers
Performance   out of the box developersPerformance   out of the box developers
Performance out of the box developersMichelle Holley
 
Supermicro and Habana High Performance, High-Efficiency AI Traning Systems
Supermicro and Habana High Performance, High-Efficiency AI Traning SystemsSupermicro and Habana High Performance, High-Efficiency AI Traning Systems
Supermicro and Habana High Performance, High-Efficiency AI Traning SystemsHabanaLabs
 
ROCm and Distributed Deep Learning on Spark and TensorFlow
ROCm and Distributed Deep Learning on Spark and TensorFlowROCm and Distributed Deep Learning on Spark and TensorFlow
ROCm and Distributed Deep Learning on Spark and TensorFlowDatabricks
 
Accelerate Your Apache Spark with Intel Optane DC Persistent Memory
Accelerate Your Apache Spark with Intel Optane DC Persistent MemoryAccelerate Your Apache Spark with Intel Optane DC Persistent Memory
Accelerate Your Apache Spark with Intel Optane DC Persistent MemoryDatabricks
 
Seyer June06 Analyst Day
Seyer June06 Analyst DaySeyer June06 Analyst Day
Seyer June06 Analyst Daylalowder
 

More Related Content

Similar to PCCC23:日本AMD株式会社 テーマ2「AMD EPYC™ プロセッサーを用いたAIソリューション」

Ibm Power System E850 pod03108 usen
Ibm Power System E850  pod03108 usenIbm Power System E850  pod03108 usen
Ibm Power System E850 pod03108 usenDiego Alberto Tamayo
 
Accelerating Virtual Machine Access with the Storage Performance Development ...
Accelerating Virtual Machine Access with the Storage Performance Development ...Accelerating Virtual Machine Access with the Storage Performance Development ...
Accelerating Virtual Machine Access with the Storage Performance Development ...Michelle Holley
 
NVMe_Infrastructure_final1.pdf
NVMe_Infrastructure_final1.pdfNVMe_Infrastructure_final1.pdf
NVMe_Infrastructure_final1.pdfIrfanBroadband
 
IBM System Storage TS7650G ProtecTIER Deduplication Gateway
IBM System Storage TS7650G ProtecTIER Deduplication GatewayIBM System Storage TS7650G ProtecTIER Deduplication Gateway
IBM System Storage TS7650G ProtecTIER Deduplication GatewayIBM India Smarter Computing
 
Dell NVIDIA AI Powered Transformation Webinar
Dell NVIDIA AI Powered Transformation WebinarDell NVIDIA AI Powered Transformation Webinar
Dell NVIDIA AI Powered Transformation WebinarBill Wong
 
DUG'20: 11 - Platform Performance Evolution from bring-up to reaching link sa...
DUG'20: 11 - Platform Performance Evolution from bring-up to reaching link sa...DUG'20: 11 - Platform Performance Evolution from bring-up to reaching link sa...
DUG'20: 11 - Platform Performance Evolution from bring-up to reaching link sa...Andrey Kudryavtsev
 
High Memory Bandwidth Demo @ One Intel Station
High Memory Bandwidth Demo @ One Intel StationHigh Memory Bandwidth Demo @ One Intel Station
High Memory Bandwidth Demo @ One Intel StationIntel IT Center
 
Computer Systems And Networks Configuration
Computer Systems And Networks ConfigurationComputer Systems And Networks Configuration
Computer Systems And Networks ConfigurationTara Daly
 
Deep Dive On Intel Optane SSDs And New Server Platforms
Deep Dive On Intel Optane SSDs And New Server PlatformsDeep Dive On Intel Optane SSDs And New Server Platforms
Deep Dive On Intel Optane SSDs And New Server PlatformsNEXTtour
 
Supermicro’s Universal GPU: Modular, Standards Based and Built for the Future
Supermicro’s Universal GPU: Modular, Standards Based and Built for the FutureSupermicro’s Universal GPU: Modular, Standards Based and Built for the Future
Supermicro’s Universal GPU: Modular, Standards Based and Built for the FutureRebekah Rodriguez
 
AMD Opteron 6000 Series Platform Press Presentation
AMD Opteron 6000 Series Platform Press PresentationAMD Opteron 6000 Series Platform Press Presentation
AMD Opteron 6000 Series Platform Press PresentationAMD
 
Re-Imagining the Data Center with Intel
Re-Imagining the Data Center with IntelRe-Imagining the Data Center with Intel
Re-Imagining the Data Center with IntelIntel IT Center
 
hbaseconasia2019 HBase Bucket Cache on Persistent Memory
hbaseconasia2019 HBase Bucket Cache on Persistent Memoryhbaseconasia2019 HBase Bucket Cache on Persistent Memory
hbaseconasia2019 HBase Bucket Cache on Persistent MemoryMichael Stack
 
Enabling ARM® Server Technology for the Datacenter
Enabling ARM® Server Technology for the DatacenterEnabling ARM® Server Technology for the Datacenter
Enabling ARM® Server Technology for the DatacenterAMD
 
TDC2019 Intel Software Day - Tecnicas de Programacao Paralela em Machine Lear...
TDC2019 Intel Software Day - Tecnicas de Programacao Paralela em Machine Lear...TDC2019 Intel Software Day - Tecnicas de Programacao Paralela em Machine Lear...
TDC2019 Intel Software Day - Tecnicas de Programacao Paralela em Machine Lear...tdc-globalcode
 
Innovation with ai at scale on the edge vt sept 2019 v0
Innovation with ai at scale  on the edge vt sept 2019 v0Innovation with ai at scale  on the edge vt sept 2019 v0
Innovation with ai at scale on the edge vt sept 2019 v0Ganesan Narayanasamy
 
Extend HPC Workloads to Amazon EC2 Instances with Intel and Rescale (CMP373-S...
Extend HPC Workloads to Amazon EC2 Instances with Intel and Rescale (CMP373-S...Extend HPC Workloads to Amazon EC2 Instances with Intel and Rescale (CMP373-S...
Extend HPC Workloads to Amazon EC2 Instances with Intel and Rescale (CMP373-S...Amazon Web Services
 

Similar to PCCC23:日本AMD株式会社 テーマ2「AMD EPYC™ プロセッサーを用いたAIソリューション」 (20)

Ibm Power System E850 pod03108 usen
Ibm Power System E850  pod03108 usenIbm Power System E850  pod03108 usen
Ibm Power System E850 pod03108 usen
 
Accelerating Virtual Machine Access with the Storage Performance Development ...
Accelerating Virtual Machine Access with the Storage Performance Development ...Accelerating Virtual Machine Access with the Storage Performance Development ...
Accelerating Virtual Machine Access with the Storage Performance Development ...
 
Amax Gpu Hpc
Amax Gpu HpcAmax Gpu Hpc
Amax Gpu Hpc
 
NVMe_Infrastructure_final1.pdf
NVMe_Infrastructure_final1.pdfNVMe_Infrastructure_final1.pdf
NVMe_Infrastructure_final1.pdf
 
IBM System Storage TS7650G ProtecTIER Deduplication Gateway
IBM System Storage TS7650G ProtecTIER Deduplication GatewayIBM System Storage TS7650G ProtecTIER Deduplication Gateway
IBM System Storage TS7650G ProtecTIER Deduplication Gateway
 
Dell NVIDIA AI Powered Transformation Webinar
Dell NVIDIA AI Powered Transformation WebinarDell NVIDIA AI Powered Transformation Webinar
Dell NVIDIA AI Powered Transformation Webinar
 
DUG'20: 11 - Platform Performance Evolution from bring-up to reaching link sa...
DUG'20: 11 - Platform Performance Evolution from bring-up to reaching link sa...DUG'20: 11 - Platform Performance Evolution from bring-up to reaching link sa...
DUG'20: 11 - Platform Performance Evolution from bring-up to reaching link sa...
 
High Memory Bandwidth Demo @ One Intel Station
High Memory Bandwidth Demo @ One Intel StationHigh Memory Bandwidth Demo @ One Intel Station
High Memory Bandwidth Demo @ One Intel Station
 
Computer Systems And Networks Configuration
Computer Systems And Networks ConfigurationComputer Systems And Networks Configuration
Computer Systems And Networks Configuration
 
Deep Dive On Intel Optane SSDs And New Server Platforms
Deep Dive On Intel Optane SSDs And New Server PlatformsDeep Dive On Intel Optane SSDs And New Server Platforms
Deep Dive On Intel Optane SSDs And New Server Platforms
 
Supermicro’s Universal GPU: Modular, Standards Based and Built for the Future
Supermicro’s Universal GPU: Modular, Standards Based and Built for the FutureSupermicro’s Universal GPU: Modular, Standards Based and Built for the Future
Supermicro’s Universal GPU: Modular, Standards Based and Built for the Future
 
Ibm power 824
Ibm power 824Ibm power 824
Ibm power 824
 
Chipsets amd
Chipsets amdChipsets amd
Chipsets amd
 
AMD Opteron 6000 Series Platform Press Presentation
AMD Opteron 6000 Series Platform Press PresentationAMD Opteron 6000 Series Platform Press Presentation
AMD Opteron 6000 Series Platform Press Presentation
 
Re-Imagining the Data Center with Intel
Re-Imagining the Data Center with IntelRe-Imagining the Data Center with Intel
Re-Imagining the Data Center with Intel
 
hbaseconasia2019 HBase Bucket Cache on Persistent Memory
hbaseconasia2019 HBase Bucket Cache on Persistent Memoryhbaseconasia2019 HBase Bucket Cache on Persistent Memory
hbaseconasia2019 HBase Bucket Cache on Persistent Memory
 
Enabling ARM® Server Technology for the Datacenter
Enabling ARM® Server Technology for the DatacenterEnabling ARM® Server Technology for the Datacenter
Enabling ARM® Server Technology for the Datacenter
 
TDC2019 Intel Software Day - Tecnicas de Programacao Paralela em Machine Lear...
TDC2019 Intel Software Day - Tecnicas de Programacao Paralela em Machine Lear...TDC2019 Intel Software Day - Tecnicas de Programacao Paralela em Machine Lear...
TDC2019 Intel Software Day - Tecnicas de Programacao Paralela em Machine Lear...
 
Innovation with ai at scale on the edge vt sept 2019 v0
Innovation with ai at scale  on the edge vt sept 2019 v0Innovation with ai at scale  on the edge vt sept 2019 v0
Innovation with ai at scale on the edge vt sept 2019 v0
 
Extend HPC Workloads to Amazon EC2 Instances with Intel and Rescale (CMP373-S...
Extend HPC Workloads to Amazon EC2 Instances with Intel and Rescale (CMP373-S...Extend HPC Workloads to Amazon EC2 Instances with Intel and Rescale (CMP373-S...
Extend HPC Workloads to Amazon EC2 Instances with Intel and Rescale (CMP373-S...
 

More from PC Cluster Consortium

PCCC23:富士通株式会社 テーマ1「次世代高性能・省電力プロセッサ『FUJITSU-MONAKA』」
PCCC23:富士通株式会社 テーマ1「次世代高性能・省電力プロセッサ『FUJITSU-MONAKA』」PCCC23:富士通株式会社 テーマ1「次世代高性能・省電力プロセッサ『FUJITSU-MONAKA』」
PCCC23:富士通株式会社 テーマ1「次世代高性能・省電力プロセッサ『FUJITSU-MONAKA』」PC Cluster Consortium
 
PCCC23:東京大学情報基盤センター 「Society5.0の実現を目指す『計算・データ・学習』の融合による革新的スーパーコンピューティング」
PCCC23:東京大学情報基盤センター 「Society5.0の実現を目指す『計算・データ・学習』の融合による革新的スーパーコンピューティング」PCCC23:東京大学情報基盤センター 「Society5.0の実現を目指す『計算・データ・学習』の融合による革新的スーパーコンピューティング」
PCCC23:東京大学情報基盤センター 「Society5.0の実現を目指す『計算・データ・学習』の融合による革新的スーパーコンピューティング」PC Cluster Consortium
 
PCCC23:日本AMD株式会社 テーマ1「AMD Instinct™ アクセラレーターの概要」
PCCC23:日本AMD株式会社 テーマ1「AMD Instinct™ アクセラレーターの概要」PCCC23:日本AMD株式会社 テーマ1「AMD Instinct™ アクセラレーターの概要」
PCCC23:日本AMD株式会社 テーマ1「AMD Instinct™ アクセラレーターの概要」PC Cluster Consortium
 
PCCC23:日本オラクル株式会社 テーマ1「OCIのHPC基盤技術と生成AI」
PCCC23:日本オラクル株式会社 テーマ1「OCIのHPC基盤技術と生成AI」PCCC23:日本オラクル株式会社 テーマ1「OCIのHPC基盤技術と生成AI」
PCCC23:日本オラクル株式会社 テーマ1「OCIのHPC基盤技術と生成AI」PC Cluster Consortium
 
PCCC23:筑波大学計算科学研究センター テーマ1「スーパーコンピュータCygnus / Pegasus」
PCCC23:筑波大学計算科学研究センター テーマ1「スーパーコンピュータCygnus / Pegasus」PCCC23:筑波大学計算科学研究センター テーマ1「スーパーコンピュータCygnus / Pegasus」
PCCC23:筑波大学計算科学研究センター テーマ1「スーパーコンピュータCygnus / Pegasus」PC Cluster Consortium
 
PCCC22:株式会社アックス テーマ1「俺ASICとロボットと論理推論AI」
PCCC22:株式会社アックス テーマ1「俺ASICとロボットと論理推論AI」PCCC22:株式会社アックス テーマ1「俺ASICとロボットと論理推論AI」
PCCC22:株式会社アックス テーマ1「俺ASICとロボットと論理推論AI」PC Cluster Consortium
 
PCCC22:日本AMD株式会社 テーマ1「第4世代AMD EPYC™ プロセッサー (Genoa) の概要」
PCCC22:日本AMD株式会社 テーマ1「第4世代AMD EPYC™ プロセッサー (Genoa) の概要」PCCC22:日本AMD株式会社 テーマ1「第4世代AMD EPYC™ プロセッサー (Genoa) の概要」
PCCC22:日本AMD株式会社 テーマ1「第4世代AMD EPYC™ プロセッサー (Genoa) の概要」PC Cluster Consortium
 
PCCC22:富士通株式会社 テーマ3「量子シミュレータ」
PCCC22:富士通株式会社 テーマ3「量子シミュレータ」PCCC22:富士通株式会社 テーマ3「量子シミュレータ」
PCCC22:富士通株式会社 テーマ3「量子シミュレータ」PC Cluster Consortium
 
PCCC22:日本電気株式会社 テーマ1「AI/ビッグデータ分析に最適なプラットフォーム NECのベクトルプロセッサ『SX-Aurora TSUBASA』」
PCCC22:日本電気株式会社 テーマ1「AI/ビッグデータ分析に最適なプラットフォーム NECのベクトルプロセッサ『SX-Aurora TSUBASA』」PCCC22:日本電気株式会社 テーマ1「AI/ビッグデータ分析に最適なプラットフォーム NECのベクトルプロセッサ『SX-Aurora TSUBASA』」
PCCC22:日本電気株式会社 テーマ1「AI/ビッグデータ分析に最適なプラットフォーム NECのベクトルプロセッサ『SX-Aurora TSUBASA』」PC Cluster Consortium
 
PCCC22:東京大学情報基盤センター 「Society5.0の実現を目指す「計算・データ・学習」の融合による革新的スーパーコンピューティング」
PCCC22:東京大学情報基盤センター 「Society5.0の実現を目指す「計算・データ・学習」の融合による革新的スーパーコンピューティング」PCCC22:東京大学情報基盤センター 「Society5.0の実現を目指す「計算・データ・学習」の融合による革新的スーパーコンピューティング」
PCCC22:東京大学情報基盤センター 「Society5.0の実現を目指す「計算・データ・学習」の融合による革新的スーパーコンピューティング」PC Cluster Consortium
 
PCCC22:日本マイクロソフト株式会社 テーマ2「HPC on Azureのお客様事例」03
PCCC22:日本マイクロソフト株式会社 テーマ2「HPC on Azureのお客様事例」03PCCC22:日本マイクロソフト株式会社 テーマ2「HPC on Azureのお客様事例」03
PCCC22:日本マイクロソフト株式会社 テーマ2「HPC on Azureのお客様事例」03PC Cluster Consortium
 
PCCC22:日本マイクロソフト株式会社 テーマ2「HPC on Azureのお客様事例」01
PCCC22:日本マイクロソフト株式会社 テーマ2「HPC on Azureのお客様事例」01PCCC22:日本マイクロソフト株式会社 テーマ2「HPC on Azureのお客様事例」01
PCCC22:日本マイクロソフト株式会社 テーマ2「HPC on Azureのお客様事例」01PC Cluster Consortium
 
PCCC22:インテル株式会社 テーマ2「次世代インテル® Xeon™ プロセッサーを中心としたインテルのHPC-AI最新情報」
PCCC22:インテル株式会社 テーマ2「次世代インテル® Xeon™ プロセッサーを中心としたインテルのHPC-AI最新情報」PCCC22:インテル株式会社 テーマ2「次世代インテル® Xeon™ プロセッサーを中心としたインテルのHPC-AI最新情報」
PCCC22:インテル株式会社 テーマ2「次世代インテル® Xeon™ プロセッサーを中心としたインテルのHPC-AI最新情報」PC Cluster Consortium
 
PCCC22:インテル株式会社 テーマ1「インテル® Agilex™ FPGA デバイス 最新情報」
PCCC22:インテル株式会社 テーマ1「インテル® Agilex™ FPGA デバイス 最新情報」PCCC22:インテル株式会社 テーマ1「インテル® Agilex™ FPGA デバイス 最新情報」
PCCC22:インテル株式会社 テーマ1「インテル® Agilex™ FPGA デバイス 最新情報」PC Cluster Consortium
 
PCCC22:筑波大学計算科学研究センター テーマ2「学際計算科学による最新の研究成果」
PCCC22:筑波大学計算科学研究センター テーマ2「学際計算科学による最新の研究成果」PCCC22:筑波大学計算科学研究センター テーマ2「学際計算科学による最新の研究成果」
PCCC22:筑波大学計算科学研究センター テーマ2「学際計算科学による最新の研究成果」PC Cluster Consortium
 
PCCC22:日本オラクル株式会社 テーマ1「Oracle Cloud Infrastructure for HPC&AI」
PCCC22:日本オラクル株式会社 テーマ1「Oracle Cloud Infrastructure for HPC&AI」PCCC22:日本オラクル株式会社 テーマ1「Oracle Cloud Infrastructure for HPC&AI」
PCCC22:日本オラクル株式会社 テーマ1「Oracle Cloud Infrastructure for HPC&AI」PC Cluster Consortium
 
PCCC21:株式会社アックス「ハイブリッドAI”ごまめ”」
PCCC21:株式会社アックス「ハイブリッドAI”ごまめ”」PCCC21:株式会社アックス「ハイブリッドAI”ごまめ”」
PCCC21:株式会社アックス「ハイブリッドAI”ごまめ”」PC Cluster Consortium
 
PCCC21:株式会社アックス「俺のハードウェア」
PCCC21:株式会社アックス「俺のハードウェア」PCCC21:株式会社アックス「俺のハードウェア」
PCCC21:株式会社アックス「俺のハードウェア」PC Cluster Consortium
 
PCCC21:東京大学情報基盤センター 「mdx: データ活用社会に向けた産学官連携のための共創プラットフォーム」
PCCC21:東京大学情報基盤センター 「mdx: データ活用社会に向けた産学官連携のための共創プラットフォーム」PCCC21:東京大学情報基盤センター 「mdx: データ活用社会に向けた産学官連携のための共創プラットフォーム」
PCCC21:東京大学情報基盤センター 「mdx: データ活用社会に向けた産学官連携のための共創プラットフォーム」PC Cluster Consortium
 
PCCC21:東京大学情報基盤センター 「『計算・データ・学習』融合によるスーパーコンピューティングの革新、そして東大センターのこれから」
PCCC21:東京大学情報基盤センター 「『計算・データ・学習』融合によるスーパーコンピューティングの革新、そして東大センターのこれから」PCCC21:東京大学情報基盤センター 「『計算・データ・学習』融合によるスーパーコンピューティングの革新、そして東大センターのこれから」
PCCC21:東京大学情報基盤センター 「『計算・データ・学習』融合によるスーパーコンピューティングの革新、そして東大センターのこれから」PC Cluster Consortium
 

More from PC Cluster Consortium (20)

PCCC23:富士通株式会社 テーマ1「次世代高性能・省電力プロセッサ『FUJITSU-MONAKA』」
PCCC23:富士通株式会社 テーマ1「次世代高性能・省電力プロセッサ『FUJITSU-MONAKA』」PCCC23:富士通株式会社 テーマ1「次世代高性能・省電力プロセッサ『FUJITSU-MONAKA』」
PCCC23:富士通株式会社 テーマ1「次世代高性能・省電力プロセッサ『FUJITSU-MONAKA』」
 
PCCC23:東京大学情報基盤センター 「Society5.0の実現を目指す『計算・データ・学習』の融合による革新的スーパーコンピューティング」
PCCC23:東京大学情報基盤センター 「Society5.0の実現を目指す『計算・データ・学習』の融合による革新的スーパーコンピューティング」PCCC23:東京大学情報基盤センター 「Society5.0の実現を目指す『計算・データ・学習』の融合による革新的スーパーコンピューティング」
PCCC23:東京大学情報基盤センター 「Society5.0の実現を目指す『計算・データ・学習』の融合による革新的スーパーコンピューティング」
 
PCCC23:日本AMD株式会社 テーマ1「AMD Instinct™ アクセラレーターの概要」
PCCC23:日本AMD株式会社 テーマ1「AMD Instinct™ アクセラレーターの概要」PCCC23:日本AMD株式会社 テーマ1「AMD Instinct™ アクセラレーターの概要」
PCCC23:日本AMD株式会社 テーマ1「AMD Instinct™ アクセラレーターの概要」
 
PCCC23:日本オラクル株式会社 テーマ1「OCIのHPC基盤技術と生成AI」
PCCC23:日本オラクル株式会社 テーマ1「OCIのHPC基盤技術と生成AI」PCCC23:日本オラクル株式会社 テーマ1「OCIのHPC基盤技術と生成AI」
PCCC23:日本オラクル株式会社 テーマ1「OCIのHPC基盤技術と生成AI」
 
PCCC23:筑波大学計算科学研究センター テーマ1「スーパーコンピュータCygnus / Pegasus」
PCCC23:筑波大学計算科学研究センター テーマ1「スーパーコンピュータCygnus / Pegasus」PCCC23:筑波大学計算科学研究センター テーマ1「スーパーコンピュータCygnus / Pegasus」
PCCC23:筑波大学計算科学研究センター テーマ1「スーパーコンピュータCygnus / Pegasus」
 
PCCC22:株式会社アックス テーマ1「俺ASICとロボットと論理推論AI」
PCCC22:株式会社アックス テーマ1「俺ASICとロボットと論理推論AI」PCCC22:株式会社アックス テーマ1「俺ASICとロボットと論理推論AI」
PCCC22:株式会社アックス テーマ1「俺ASICとロボットと論理推論AI」
 
PCCC22:日本AMD株式会社 テーマ1「第4世代AMD EPYC™ プロセッサー (Genoa) の概要」
PCCC22:日本AMD株式会社 テーマ1「第4世代AMD EPYC™ プロセッサー (Genoa) の概要」PCCC22:日本AMD株式会社 テーマ1「第4世代AMD EPYC™ プロセッサー (Genoa) の概要」
PCCC22:日本AMD株式会社 テーマ1「第4世代AMD EPYC™ プロセッサー (Genoa) の概要」
 
PCCC22:富士通株式会社 テーマ3「量子シミュレータ」
PCCC22:富士通株式会社 テーマ3「量子シミュレータ」PCCC22:富士通株式会社 テーマ3「量子シミュレータ」
PCCC22:富士通株式会社 テーマ3「量子シミュレータ」
 
PCCC22:日本電気株式会社 テーマ1「AI/ビッグデータ分析に最適なプラットフォーム NECのベクトルプロセッサ『SX-Aurora TSUBASA』」
PCCC22:日本電気株式会社 テーマ1「AI/ビッグデータ分析に最適なプラットフォーム NECのベクトルプロセッサ『SX-Aurora TSUBASA』」PCCC22:日本電気株式会社 テーマ1「AI/ビッグデータ分析に最適なプラットフォーム NECのベクトルプロセッサ『SX-Aurora TSUBASA』」
PCCC22:日本電気株式会社 テーマ1「AI/ビッグデータ分析に最適なプラットフォーム NECのベクトルプロセッサ『SX-Aurora TSUBASA』」
 
PCCC22:東京大学情報基盤センター 「Society5.0の実現を目指す「計算・データ・学習」の融合による革新的スーパーコンピューティング」
PCCC22:東京大学情報基盤センター 「Society5.0の実現を目指す「計算・データ・学習」の融合による革新的スーパーコンピューティング」PCCC22:東京大学情報基盤センター 「Society5.0の実現を目指す「計算・データ・学習」の融合による革新的スーパーコンピューティング」
PCCC22:東京大学情報基盤センター 「Society5.0の実現を目指す「計算・データ・学習」の融合による革新的スーパーコンピューティング」
 
PCCC22:日本マイクロソフト株式会社 テーマ2「HPC on Azureのお客様事例」03
PCCC22:日本マイクロソフト株式会社 テーマ2「HPC on Azureのお客様事例」03PCCC22:日本マイクロソフト株式会社 テーマ2「HPC on Azureのお客様事例」03
PCCC22:日本マイクロソフト株式会社 テーマ2「HPC on Azureのお客様事例」03
 
PCCC22:日本マイクロソフト株式会社 テーマ2「HPC on Azureのお客様事例」01
PCCC22:日本マイクロソフト株式会社 テーマ2「HPC on Azureのお客様事例」01PCCC22:日本マイクロソフト株式会社 テーマ2「HPC on Azureのお客様事例」01
PCCC22:日本マイクロソフト株式会社 テーマ2「HPC on Azureのお客様事例」01
 
PCCC22:インテル株式会社 テーマ2「次世代インテル® Xeon™ プロセッサーを中心としたインテルのHPC-AI最新情報」
PCCC22:インテル株式会社 テーマ2「次世代インテル® Xeon™ プロセッサーを中心としたインテルのHPC-AI最新情報」PCCC22:インテル株式会社 テーマ2「次世代インテル® Xeon™ プロセッサーを中心としたインテルのHPC-AI最新情報」
PCCC22:インテル株式会社 テーマ2「次世代インテル® Xeon™ プロセッサーを中心としたインテルのHPC-AI最新情報」
 
PCCC22:インテル株式会社 テーマ1「インテル® Agilex™ FPGA デバイス 最新情報」
PCCC22:インテル株式会社 テーマ1「インテル® Agilex™ FPGA デバイス 最新情報」PCCC22:インテル株式会社 テーマ1「インテル® Agilex™ FPGA デバイス 最新情報」
PCCC22:インテル株式会社 テーマ1「インテル® Agilex™ FPGA デバイス 最新情報」
 
PCCC22:筑波大学計算科学研究センター テーマ2「学際計算科学による最新の研究成果」
PCCC22:筑波大学計算科学研究センター テーマ2「学際計算科学による最新の研究成果」PCCC22:筑波大学計算科学研究センター テーマ2「学際計算科学による最新の研究成果」
PCCC22:筑波大学計算科学研究センター テーマ2「学際計算科学による最新の研究成果」
 
PCCC22:日本オラクル株式会社 テーマ1「Oracle Cloud Infrastructure for HPC&AI」
PCCC22:日本オラクル株式会社 テーマ1「Oracle Cloud Infrastructure for HPC&AI」PCCC22:日本オラクル株式会社 テーマ1「Oracle Cloud Infrastructure for HPC&AI」
PCCC22:日本オラクル株式会社 テーマ1「Oracle Cloud Infrastructure for HPC&AI」
 
PCCC21:株式会社アックス「ハイブリッドAI”ごまめ”」
PCCC21:株式会社アックス「ハイブリッドAI”ごまめ”」PCCC21:株式会社アックス「ハイブリッドAI”ごまめ”」
PCCC21:株式会社アックス「ハイブリッドAI”ごまめ”」
 
PCCC21:株式会社アックス「俺のハードウェア」
PCCC21:株式会社アックス「俺のハードウェア」PCCC21:株式会社アックス「俺のハードウェア」
PCCC21:株式会社アックス「俺のハードウェア」
 
PCCC21:東京大学情報基盤センター 「mdx: データ活用社会に向けた産学官連携のための共創プラットフォーム」
PCCC21:東京大学情報基盤センター 「mdx: データ活用社会に向けた産学官連携のための共創プラットフォーム」PCCC21:東京大学情報基盤センター 「mdx: データ活用社会に向けた産学官連携のための共創プラットフォーム」
PCCC21:東京大学情報基盤センター 「mdx: データ活用社会に向けた産学官連携のための共創プラットフォーム」
 
PCCC21:東京大学情報基盤センター 「『計算・データ・学習』融合によるスーパーコンピューティングの革新、そして東大センターのこれから」
PCCC21:東京大学情報基盤センター 「『計算・データ・学習』融合によるスーパーコンピューティングの革新、そして東大センターのこれから」PCCC21:東京大学情報基盤センター 「『計算・データ・学習』融合によるスーパーコンピューティングの革新、そして東大センターのこれから」
PCCC21:東京大学情報基盤センター 「『計算・データ・学習』融合によるスーパーコンピューティングの革新、そして東大センターのこれから」
 

Recently uploaded

"The Transformative Power of AI and Open Challenges" by Dr. Manish Gupta, Google
"The Transformative Power of AI and Open Challenges" by Dr. Manish Gupta, Google"The Transformative Power of AI and Open Challenges" by Dr. Manish Gupta, Google
"The Transformative Power of AI and Open Challenges" by Dr. Manish Gupta, GoogleISPMAIndia
 
"Running Open-Source LLM models on Kubernetes", Volodymyr Tsap
"Running Open-Source LLM models on Kubernetes",  Volodymyr Tsap"Running Open-Source LLM models on Kubernetes",  Volodymyr Tsap
"Running Open-Source LLM models on Kubernetes", Volodymyr TsapFwdays
 
Battle of React State Managers in frontend applications
Battle of React State Managers in frontend applicationsBattle of React State Managers in frontend applications
Battle of React State Managers in frontend applicationsEvangelia Mitsopoulou
 
"AIRe - AI Reliability Engineering", Denys Vasyliev
"AIRe - AI Reliability Engineering", Denys Vasyliev"AIRe - AI Reliability Engineering", Denys Vasyliev
"AIRe - AI Reliability Engineering", Denys VasylievFwdays
 
The Art of the Possible with Graph by Dr Jim Webber Neo4j.pptx
The Art of the Possible with Graph by Dr Jim Webber Neo4j.pptxThe Art of the Possible with Graph by Dr Jim Webber Neo4j.pptx
The Art of the Possible with Graph by Dr Jim Webber Neo4j.pptxNeo4j
 
Imaging and Design for the Online Environment Part 1.pptx
Imaging and Design for the Online Environment Part 1.pptxImaging and Design for the Online Environment Part 1.pptx
Imaging and Design for the Online Environment Part 1.pptxPower Point
 
My Journey towards Artificial Intelligence
My Journey towards Artificial IntelligenceMy Journey towards Artificial Intelligence
My Journey towards Artificial IntelligenceVijayananda Mohire
 
Progress Report: Ministry of IT under Dr. Umar Saif Aug 23-Feb'24
Progress Report: Ministry of IT under Dr. Umar Saif Aug 23-Feb'24Progress Report: Ministry of IT under Dr. Umar Saif Aug 23-Feb'24
Progress Report: Ministry of IT under Dr. Umar Saif Aug 23-Feb'24Umar Saif
 
Digital Transformation Strategy & Plan Templates - www.beyondthecloud.digital...
Digital Transformation Strategy & Plan Templates - www.beyondthecloud.digital...Digital Transformation Strategy & Plan Templates - www.beyondthecloud.digital...
Digital Transformation Strategy & Plan Templates - www.beyondthecloud.digital...MarcovanHurne2
 
IT Nation Evolve event 2024 - Quarter 1
IT Nation Evolve event 2024  - Quarter 1IT Nation Evolve event 2024  - Quarter 1
IT Nation Evolve event 2024 - Quarter 1Inbay UK
 
21ST CENTURY LITERACY FROM TRADITIONAL TO MODERN
21ST CENTURY LITERACY FROM TRADITIONAL TO MODERN21ST CENTURY LITERACY FROM TRADITIONAL TO MODERN
21ST CENTURY LITERACY FROM TRADITIONAL TO MODERNRonnelBaroc
 
Act Like an Owner, Challenge Like a VC by former CPO, Tripadvisor
Act Like an Owner,  Challenge Like a VC by former CPO, TripadvisorAct Like an Owner,  Challenge Like a VC by former CPO, Tripadvisor
Act Like an Owner, Challenge Like a VC by former CPO, TripadvisorProduct School
 
The Future of Product, by Founder & CEO, Product School
The Future of Product, by Founder & CEO, Product SchoolThe Future of Product, by Founder & CEO, Product School
The Future of Product, by Founder & CEO, Product SchoolProduct School
 
Dynamical systems simulation in Python for science and engineering
Dynamical systems simulation in Python for science and engineeringDynamical systems simulation in Python for science and engineering
Dynamical systems simulation in Python for science and engineeringMassimo Talia
 
Leveraging SLF4j for Effective Logging in IBM App Connect Enterprise.docx
Leveraging SLF4j for Effective Logging in IBM App Connect Enterprise.docxLeveraging SLF4j for Effective Logging in IBM App Connect Enterprise.docx
Leveraging SLF4j for Effective Logging in IBM App Connect Enterprise.docxVotarikari Shravan
 
AI Act & Standardization: UNINFO involvement
AI Act & Standardization: UNINFO involvementAI Act & Standardization: UNINFO involvement
AI Act & Standardization: UNINFO involvementMimmo Squillace
 
Cultivating Entrepreneurial Mindset in Product Management: Strategies for Suc...
Cultivating Entrepreneurial Mindset in Product Management: Strategies for Suc...Cultivating Entrepreneurial Mindset in Product Management: Strategies for Suc...
Cultivating Entrepreneurial Mindset in Product Management: Strategies for Suc...Product School
 
"DevOps Practisting Platform on EKS with Karpenter autoscaling", Dmytro Kozhevin
"DevOps Practisting Platform on EKS with Karpenter autoscaling", Dmytro Kozhevin"DevOps Practisting Platform on EKS with Karpenter autoscaling", Dmytro Kozhevin
"DevOps Practisting Platform on EKS with Karpenter autoscaling", Dmytro KozhevinFwdays
 
Automation Ops Series: Session 1 - Introduction and setup DevOps for UiPath p...
Automation Ops Series: Session 1 - Introduction and setup DevOps for UiPath p...Automation Ops Series: Session 1 - Introduction and setup DevOps for UiPath p...
Automation Ops Series: Session 1 - Introduction and setup DevOps for UiPath p...DianaGray10
 
How to write an effective Cyber Incident Response Plan
How to write an effective Cyber Incident Response PlanHow to write an effective Cyber Incident Response Plan
How to write an effective Cyber Incident Response PlanDatabarracks
 

Recently uploaded (20)

"The Transformative Power of AI and Open Challenges" by Dr. Manish Gupta, Google
"The Transformative Power of AI and Open Challenges" by Dr. Manish Gupta, Google"The Transformative Power of AI and Open Challenges" by Dr. Manish Gupta, Google
"The Transformative Power of AI and Open Challenges" by Dr. Manish Gupta, Google
 
"Running Open-Source LLM models on Kubernetes", Volodymyr Tsap
"Running Open-Source LLM models on Kubernetes",  Volodymyr Tsap"Running Open-Source LLM models on Kubernetes",  Volodymyr Tsap
"Running Open-Source LLM models on Kubernetes", Volodymyr Tsap
 
Battle of React State Managers in frontend applications
Battle of React State Managers in frontend applicationsBattle of React State Managers in frontend applications
Battle of React State Managers in frontend applications
 
"AIRe - AI Reliability Engineering", Denys Vasyliev
"AIRe - AI Reliability Engineering", Denys Vasyliev"AIRe - AI Reliability Engineering", Denys Vasyliev
"AIRe - AI Reliability Engineering", Denys Vasyliev
 
The Art of the Possible with Graph by Dr Jim Webber Neo4j.pptx
The Art of the Possible with Graph by Dr Jim Webber Neo4j.pptxThe Art of the Possible with Graph by Dr Jim Webber Neo4j.pptx
The Art of the Possible with Graph by Dr Jim Webber Neo4j.pptx
 
Imaging and Design for the Online Environment Part 1.pptx
Imaging and Design for the Online Environment Part 1.pptxImaging and Design for the Online Environment Part 1.pptx
Imaging and Design for the Online Environment Part 1.pptx
 
My Journey towards Artificial Intelligence
My Journey towards Artificial IntelligenceMy Journey towards Artificial Intelligence
My Journey towards Artificial Intelligence
 
Progress Report: Ministry of IT under Dr. Umar Saif Aug 23-Feb'24
Progress Report: Ministry of IT under Dr. Umar Saif Aug 23-Feb'24Progress Report: Ministry of IT under Dr. Umar Saif Aug 23-Feb'24
Progress Report: Ministry of IT under Dr. Umar Saif Aug 23-Feb'24
 
Digital Transformation Strategy & Plan Templates - www.beyondthecloud.digital...
Digital Transformation Strategy & Plan Templates - www.beyondthecloud.digital...Digital Transformation Strategy & Plan Templates - www.beyondthecloud.digital...
Digital Transformation Strategy & Plan Templates - www.beyondthecloud.digital...
 
IT Nation Evolve event 2024 - Quarter 1
IT Nation Evolve event 2024  - Quarter 1IT Nation Evolve event 2024  - Quarter 1
IT Nation Evolve event 2024 - Quarter 1
 
21ST CENTURY LITERACY FROM TRADITIONAL TO MODERN
21ST CENTURY LITERACY FROM TRADITIONAL TO MODERN21ST CENTURY LITERACY FROM TRADITIONAL TO MODERN
21ST CENTURY LITERACY FROM TRADITIONAL TO MODERN
 
Act Like an Owner, Challenge Like a VC by former CPO, Tripadvisor
Act Like an Owner,  Challenge Like a VC by former CPO, TripadvisorAct Like an Owner,  Challenge Like a VC by former CPO, Tripadvisor
Act Like an Owner, Challenge Like a VC by former CPO, Tripadvisor
 
The Future of Product, by Founder & CEO, Product School
The Future of Product, by Founder & CEO, Product SchoolThe Future of Product, by Founder & CEO, Product School
The Future of Product, by Founder & CEO, Product School
 
Dynamical systems simulation in Python for science and engineering
Dynamical systems simulation in Python for science and engineeringDynamical systems simulation in Python for science and engineering
Dynamical systems simulation in Python for science and engineering
 
Leveraging SLF4j for Effective Logging in IBM App Connect Enterprise.docx
Leveraging SLF4j for Effective Logging in IBM App Connect Enterprise.docxLeveraging SLF4j for Effective Logging in IBM App Connect Enterprise.docx
Leveraging SLF4j for Effective Logging in IBM App Connect Enterprise.docx
 
AI Act & Standardization: UNINFO involvement
AI Act & Standardization: UNINFO involvementAI Act & Standardization: UNINFO involvement
AI Act & Standardization: UNINFO involvement
 
Cultivating Entrepreneurial Mindset in Product Management: Strategies for Suc...
Cultivating Entrepreneurial Mindset in Product Management: Strategies for Suc...Cultivating Entrepreneurial Mindset in Product Management: Strategies for Suc...
Cultivating Entrepreneurial Mindset in Product Management: Strategies for Suc...
 
"DevOps Practisting Platform on EKS with Karpenter autoscaling", Dmytro Kozhevin
"DevOps Practisting Platform on EKS with Karpenter autoscaling", Dmytro Kozhevin"DevOps Practisting Platform on EKS with Karpenter autoscaling", Dmytro Kozhevin
"DevOps Practisting Platform on EKS with Karpenter autoscaling", Dmytro Kozhevin
 
Automation Ops Series: Session 1 - Introduction and setup DevOps for UiPath p...
Automation Ops Series: Session 1 - Introduction and setup DevOps for UiPath p...Automation Ops Series: Session 1 - Introduction and setup DevOps for UiPath p...
Automation Ops Series: Session 1 - Introduction and setup DevOps for UiPath p...
 
How to write an effective Cyber Incident Response Plan
How to write an effective Cyber Incident Response PlanHow to write an effective Cyber Incident Response Plan
How to write an effective Cyber Incident Response Plan
 

PCCC23:日本AMD株式会社 テーマ2「AMD EPYC™ プロセッサーを用いたAIソリューション」

  • 2. 2 | [Public] Broad Industry Impact Self-driving cars can recognize signage, pedestrians, and other vehicles to be avoided Monitor quality of manufactured products from food items to printed circuit boards Automate checkout lines and use product recommendation engines to offer suggestions, whether online or in the store Detect anomalies including fractures and tumors. Use the same models in research to assess in vitro cell growth and proliferation Natural language processing can use spoken requests and recommendation engines to help point customers to solutions AI-powered anomaly detection helps stop credit-card fraud, while computer vision models watch for suspicious documents Automotive Manufacturing Retail Financial Services Medical Service Automation AI extends and enriches common business workloads and activities
  • 3. 3 | [Public] AMD Propels theAI Lifecycle TRAINING The most data- and processing-intensive part of the AI lifecycle. Significant computing power is required, and servers equipped with AMD Instinct™ accelerators are designed to accelerate the process INFERENCING Once trained, AI requires comparatively less processing power to process incoming data and business records in real time. Inferencing happens close to the data and AMD EPYC™ processors are ideal for inferencing.
  • 4. 4 | [Public] AMD EPYC™ Processors: Inference Performance 1.78x Up to SERVERS BASED ON AMD EPYC™ 9654 CPUS RECOGNIZE VEHICLES AT 1.78X THE RATE OF INTEL® XEON® Platinum 8940H CPU BASED SERVERS Phoronix used the OpenVINO benchmark using INT8-FP16 data types to compare multiple CPU types. They measured a whopping 78% speedup on vehicle detection, and a 14% speedup on age-gender recognition comparing a 2P Intel Xeon 8490H processor-powered server to a 2P AMD EPYC 9654 processor-powered server with ATX-512 on. See endnotes SP5-192, -193 6207 11029 0 2000 4000 6000 8000 10000 12000 2 x Intel Xeon Platinum 8940H 2 x AMD EPYC 9654 OpenVINO FP16-INT8 Vehicle Detection FPS (Higher is Better) 103184 118104 0 20000 40000 60000 80000 100000 120000 140000 2 x Intel Xeon Platinum 8940H 2 x AMD EPYC 9654 OpenVINO FP16-INT8 Age Gender Recognition Faces per Second (Higher is Better)
  • 5. 5 | [Public] 256 Threads for End-to-EndAI Boost Results may vary due to factors including system configurations, software versions and BIOS settings. As of 6/13/2023, see endnotes: SP5-051. • Comparison derived from TPCx-AI benchmark covering 10 end-to-end use cases covering training, serving and throughput • 128C AMD EPYC 9754 delivers up to an aggregate of ~2.2x the AI test cases per min. vs. 60C Intel Xeon Platinum 8490H Outstanding end-to-end AI throughput performance on a wide variety of use cases 831 1841 120 total cores/ 240 threads 4 instances / 30 vCPUs per 256 total cores/ 512 threads 8 instances / 30 vCPU per Xeon® Platinum 8490H AMD EPYC™ 9754 ~2.2x Running 2Pserverswith 128C4thGenAMDEPYC™ 9754vs.60C4thGenIntel® Xeon® Platinum8490H End-to-end AI data science pipeline aggregate AI use cases/min
  • 6. 6 | [Public] ROCm™ Platform Vitis™ AI Platform CPU Stack Unified Inferencing Model StreamlinesAdoption The Unified Inference Frontend (blue) provides a uniform way to link your inferencing software with the acceleration capabilities of EPYC™ CPUs, AMD Instinct™ accelerators, and Versal™ and Zynq™ adaptive SoCs The CPU-specific software stack includes a robust set of tools that accelerate deep learning and inference workloads
  • 7. 7 | [Public] AI Ecosystem Enablement Model Optimization • ResNet50, ResNet101, ResNet152 • MobileNet-v1, MobileNet-v2 • Inception V3, Inception V4 • AlexNet, GoogleNet • RNNs, LSTMs, GRUs • BERT-Base, BERT-Large • DLRM • Wide and Deep Key Models that use Optimized ZenDNN Primitives Computer Vision Natural Language Processing Recommendation Systems
  • 8. 8 | [Public] AMD Solutions forAI Workload-optimized engines enableAI efficiency AI Accelerators Server CPUs FPGAs and Adaptive SoCs
  • 11. 11 | [Public] End Notes SP5-051: TPCx-AI SF3 derivative workload comparison based on AMD internal testing running multiple VM instances as of 6/13/2023. The aggregate end-to-end AI throughput test is derived from the TPCx-AI benchmark and as such is not comparable to published TPCx-AI results, as the end-to-end AI throughput test results do not comply with the TPCx-AI Specification. Configurations: 2 x AMD EPYC 9754 on Titanite (BIOS and Settings: AMI Core Ver. 5.25, Project Ver. RTI1000F and Default BIOS settings (SMT=on, Determinism=Auto, NPS=1)), 1.5TB (24) Dual-Rank DDR5-4800 64GB DIMMs, 1DPC, SK Hynix SHGP31-500GM 500GB NVMe, Ubuntu® 22.04 LTS (8-instances, 30 vCPUs/instance, 1841 AI test cases/min); 2 x AMD EPYC 9654 on Titanite (BIOS and Settings: AMI Core Ver. 5.25, Project Ver. RTI1000F and Default BIOS settings (SMT=on, Determinism=Auto, NPS=1)), 1.5TB (24) Dual-Rank DDR5-4800 64GB DIMMs, 1DPC, Samsung SSD 983 DCT 960GB, Ubuntu 22.04.1 LTS (6-instance, 28 vCPUs/instance, 1554 AI test cases/min); 2 x Intel(R) Xeon(R) Platinum 8490H on Dell PowerEdge R760 (BIOS and Settings: ESE110Q-1.10 and Package C1E, Default BIOS settings (C State=Disabled, Hyper-Threading, Turbo boost= enabled (ALL)=Enabled, SNC (Sub NUMA)=Disabled)), 2TB (32) Dual-Rank DDR5-4800 64GB DIMMs, 1DPC, Dell 1.7TB NVMe, Ubuntu 22.04.2 LTS (4-instance, 30 vCPUs/instance, 831 AI test cases/min). Results may vary due to factors including system configurations, software versions and BIOS settings. TPC Benchmark is a trademark of the TPC. SP5-192: OpenVINO 2022.2 FP16-INT8 Vehicle Detection FPS with AVX-512 on comparison based on Phoronix Test as of 18 Jan 2023. Configurations: 2P 96-core AMD EPYC™ 9654 (11029 FPS) powered server versus 2P 60-core Intel® Xeon® Platinum 8940H (6207 FPS) for 1.78x the performance. https://www.phoronix.com/review/intel-sapphirerapids-avx512/7. Testing not independently verified by AMD. Scores will vary based on system configuration and determinism mode used. SP5-193: OpenVINO 2022.3 FP16-INT8 Age Gender Recognition Faces per Second comparison based on Phoronix Test as of 18 Jan 2023. Configurations: 2P 96-core AMD EPYC™ 9654 (118104 Faces per Second) powered server versus 2P 60-core Intel® Xeon® Platinum 8940H (103184 Faces per Second) for 1.14x the performance. https://www.phoronix.com/review/intel-sapphirerapids-avx512/7. Testing not independently verified by AMD. Scores will vary based on system configuration and determinism mode used. Reference “AI Inferencing with AMD EPYC Processors” : https://www.amd.com/content/dam/amd/en/documents/solutions/ai/ai-inferencing-amd-epyc-processors-white-paper.pdf
  • 12. 12 | [Public] DISCLAIMERS AND ATTRIBUTIONS The information contained herein is for informational purposes only and is subject to change without notice. While every precaution has been taken in the preparation of this document, it may contain technical inaccuracies, omissions and typographical errors, and AMD is under no obligation to update or otherwise correct this information. Advanced Micro Devices, Inc. makes no representations or warranties with respect to the accuracy or completeness of the contents of this document, and assumes no liability of any kind, including the implied warranties of noninfringement, merchantability or fitness for particular purposes, with respect to the operation or use of AMD hardware, software or other products described herein. No license, including implied or arising by estoppel, to any intellectual property rights is granted by this document. Terms and limitations applicable to the purchase or use of AMD’s products are as set forth in a signed agreement between the parties or in AMD's Standard Terms and Conditions of Sale. GD-18 ©2023 Advanced Micro Devices, Inc. all rights reserved. AMD, the AMD arrow, EPYC, and combinations thereof are trademarks of Advanced Micro Devices, Inc Intel, the Intel logo and Xeon are trademarks of Intel Corporation or its subsidiaries. SPEC®, SPECrate® and SPEC CPU® are registered trademarks of the Standard Performance Evaluation Corporation. See www.spec.org for more information. Other product names used in this publication are for identification purposes only and may be trademarks of their respective companies.