SlideShare a Scribd company logo
1 of 12
Download to read offline
Evaluation of ARM CPUs
for IceCube
available through
Google Kubernetes Engine
Igor Sfiligoi – UCSD
in collaboration with David Schultz & Benedikt Riedel - University of Wisconsin Madison
CMS O&C Meeting, 9/28/22
Benchmark setup
• Google offered some credits to IceCube to be used on the
Google Kubernetes Engine (GKE)
• A managed Kubernetes setup on the Google Compute Cloud (GCP)
• Google offers several generations of CPUs
• x86 from both Intel and AMD, in both Hyper-threaded and not mode
• ARM CPUs from Ampere Computing (no hyper-threading)
• The main CPU-heavy application is CORSIKA
• Cosmic air shower simulator
• x86 binaries available on CVMFS
• ARM binaries compiled in GKE
CMS O&C Meeting, 9/28/22
CPUs used
• Being a Cloud platform, a lot of details are hidden
• But this is what I can see:
• t2a - one vCPU per core
• Ampere Altra Arm Q64-30 processor with all-core frequency of 3.0 GHz
• t2d - one vCPU per core
• AMD EPYC 7B13 - 2.45 base 2.8 effective 3.5 max boost
• c2d - two vCPU per core
• AMD EPYC 7B13 - 2.45 base 2.8 effective 3.5 max boost
• c2 - two vCPUs per core
• Intel® Xeon® Gold 6253CL - 3.1 base 3.8 all core turbo 3.9 single core
Same HW, different setup
CMS O&C Meeting, 9/28/22
Benchmark results
ARM CPUs faster than all tested x86 CPUs
Hyper-threading actually
quite effective
ARM
CMS O&C Meeting, 9/28/22
Cost effectiveness
But AMD pricing definitely
not in line with performance
ARM CPUs more cost-effective than x86 CPUs
ARM
CMS O&C Meeting, 9/28/22
Porting experience
• x86 binaries already available, so a no brainer
• We did not have ARM binaries, so we had to compile from source
• Likely the first time it was done for most of the IceCube SW stack
• The process was mostly painless
• Some hardcoded x86-specific compiler parameters
• Easy to replace with ARM-equivalents (but took a few passes to get them all)
• CORSIKA itself built without a hitch
• But we did need a bunch of helper binaries for the jobs to run
• OS and the like came out of the box for both platforms
• As standard docker images (we used k8s, remember?)
CMS O&C Meeting, 9/28/22
This is it for concrete data
• The previous slides contain pretty much all that we have done
• In the next few slides I will show a few reference data
about power consumption I pulled from the Web
• Take with a grain of salt!
CMS O&C Meeting, 9/28/22
CPU comparison
• Ampere Altra Q64-30
• 64 cores at 3.0 GHz
• 180W TDP
• 8x 3200 DDR4 RAM
• AMD EPYC 7B13
• CPU family 25, Stepping 1
• 60 cores, base freq 2.45 Ghz
• 8x 3200 DDR4 RAM
• Closest public TDP data is for
EPYC 7713 at 225W TDP
https://www.anandtech.com/show/16979/the-ampere-
altra-max-review-pushing-it-to-128-cores-per-socket
https://www.amd.com/en/products/cpu/amd-epyc-7713
https://cloud.google.com/compute/docs/general-purpose-
machines#t2d_machines
Intel Xeon 6253CL supposedly has 18 cores at 200W TDP
CMS O&C Meeting, 9/28/22
CPU comparison
• Ampere Altra Q64-30
• 64 cores at 3.0 GHz
• 180W TDP
• 8x 3200 DDR4 RAM
• AMD EPYC 7B13
• CPU family 25, Stepping 1
• 60 cores, base freq 2.45 Ghz
• 8x 3200 DDR4 RAM
• Closest public TDP data is for
EPYC 7713 at 225W TDP
https://www.anandtech.com/show/16979/the-ampere-
altra-max-review-pushing-it-to-128-cores-per-socket
https://www.amd.com/en/products/cpu/amd-epyc-7713
https://cloud.google.com/compute/docs/general-purpose-
machines#t2d_machines
Intel Xeon 6253CL supposedly has 18 cores at 200W TDP
CPU power
consumption is just a
fraction of the total
node power usage, of
course!
CMS O&C Meeting, 9/28/22
CPU power effectiveness
AMD EPYC in HT mode likely beats ARM CPU in power effectiveness.
But ARM CPU more power efficient when non-HT mode is desired.
Hyper-threading great
for power effectiveness
Disclaimer:
Power usage
based on
estimated TDP
CMS O&C Meeting, 9/28/22
In summary
• ARM CPUs now easily available in the commercial Clouds
• Porting code to ARM not hard if you have the source code
• In GKE, using IceCube CORSIKA simulation
• ARM cores about as fast as the x86 cores (even a bit faster)
• Priced cheaper, too
• No hard data on power usage, but extrapolated data imply
• ARM CPUs only modestly more power efficient that x86 competition
CMS O&C Meeting, 9/28/22
Acknowledgement
• This work was partially funded by the U.S. National Science
Foundation (NSF) under grants OAC-1826967, OAC-1541349,
OPP- 2042807, CNS-1925001 and OAC- 2103963.
• All Google Kubernetes Engine costs have been covered by Google-
issued credits.
CMS O&C Meeting, 9/28/22

More Related Content

Similar to O&C Meeting - Evaluation of ARM CPUs for IceCube available through Google Kubernetes Engine

RMOUG 2013 - Where did my CPU go?
RMOUG 2013 - Where did my CPU go?RMOUG 2013 - Where did my CPU go?
RMOUG 2013 - Where did my CPU go?Kristofferson A
 
Heterogeneous Computing on POWER - IBM and OpenPOWER technologies to accelera...
Heterogeneous Computing on POWER - IBM and OpenPOWER technologies to accelera...Heterogeneous Computing on POWER - IBM and OpenPOWER technologies to accelera...
Heterogeneous Computing on POWER - IBM and OpenPOWER technologies to accelera...Cesar Maciel
 
FPGAs in the cloud? (October 2017)
FPGAs in the cloud? (October 2017)FPGAs in the cloud? (October 2017)
FPGAs in the cloud? (October 2017)Julien SIMON
 
Achieving the ultimate performance with KVM
Achieving the ultimate performance with KVMAchieving the ultimate performance with KVM
Achieving the ultimate performance with KVMStorPool Storage
 
Choose Your Weapon: Comparing Spark on FPGAs vs GPUs
Choose Your Weapon: Comparing Spark on FPGAs vs GPUsChoose Your Weapon: Comparing Spark on FPGAs vs GPUs
Choose Your Weapon: Comparing Spark on FPGAs vs GPUsDatabricks
 
AWS Summit Bogotá Track Avanzado: EC2 avanzado
AWS Summit Bogotá Track Avanzado: EC2 avanzadoAWS Summit Bogotá Track Avanzado: EC2 avanzado
AWS Summit Bogotá Track Avanzado: EC2 avanzadoAmazon Web Services
 
Making the most out of Heterogeneous Chips with CPU, GPU and FPGA
Making the most out of Heterogeneous Chips with CPU, GPU and FPGAMaking the most out of Heterogeneous Chips with CPU, GPU and FPGA
Making the most out of Heterogeneous Chips with CPU, GPU and FPGAFacultad de Informática UCM
 
Tackling Network Bottlenecks with Hardware Accelerations: Cloud vs. On-Premise
Tackling Network Bottlenecks with Hardware Accelerations: Cloud vs. On-PremiseTackling Network Bottlenecks with Hardware Accelerations: Cloud vs. On-Premise
Tackling Network Bottlenecks with Hardware Accelerations: Cloud vs. On-PremiseDatabricks
 
”Bare-Metal Container" presented at HPCC2016
”Bare-Metal Container" presented at HPCC2016”Bare-Metal Container" presented at HPCC2016
”Bare-Metal Container" presented at HPCC2016Kuniyasu Suzaki
 
Using GPUs to handle Big Data with Java by Adam Roberts.
Using GPUs to handle Big Data with Java by Adam Roberts.Using GPUs to handle Big Data with Java by Adam Roberts.
Using GPUs to handle Big Data with Java by Adam Roberts.J On The Beach
 
Arm Neoverse market update_05122020.pdf
Arm Neoverse market update_05122020.pdfArm Neoverse market update_05122020.pdf
Arm Neoverse market update_05122020.pdfPaul Yang
 
The Rise of Parallel Computing
The Rise of Parallel ComputingThe Rise of Parallel Computing
The Rise of Parallel Computingbakers84
 
QEMU and Raspberry Pi. Instant Embedded Development
QEMU and Raspberry Pi. Instant Embedded DevelopmentQEMU and Raspberry Pi. Instant Embedded Development
QEMU and Raspberry Pi. Instant Embedded DevelopmentGlobalLogic Ukraine
 
Deep Dive on Delivering Amazon EC2 Instance Performance
Deep Dive on Delivering Amazon EC2 Instance PerformanceDeep Dive on Delivering Amazon EC2 Instance Performance
Deep Dive on Delivering Amazon EC2 Instance PerformanceAmazon Web Services
 
Computação acelerada – a era das ap us roberto brandão, ciência
Computação acelerada – a era das ap us   roberto brandão,  ciênciaComputação acelerada – a era das ap us   roberto brandão,  ciência
Computação acelerada – a era das ap us roberto brandão, ciênciaCampus Party Brasil
 
Achieving the Ultimate Performance with KVM
Achieving the Ultimate Performance with KVMAchieving the Ultimate Performance with KVM
Achieving the Ultimate Performance with KVMDevOps.com
 

Similar to O&C Meeting - Evaluation of ARM CPUs for IceCube available through Google Kubernetes Engine (20)

RMOUG 2013 - Where did my CPU go?
RMOUG 2013 - Where did my CPU go?RMOUG 2013 - Where did my CPU go?
RMOUG 2013 - Where did my CPU go?
 
Heterogeneous Computing on POWER - IBM and OpenPOWER technologies to accelera...
Heterogeneous Computing on POWER - IBM and OpenPOWER technologies to accelera...Heterogeneous Computing on POWER - IBM and OpenPOWER technologies to accelera...
Heterogeneous Computing on POWER - IBM and OpenPOWER technologies to accelera...
 
FPGAs in the cloud? (October 2017)
FPGAs in the cloud? (October 2017)FPGAs in the cloud? (October 2017)
FPGAs in the cloud? (October 2017)
 
Achieving the ultimate performance with KVM
Achieving the ultimate performance with KVMAchieving the ultimate performance with KVM
Achieving the ultimate performance with KVM
 
Choose Your Weapon: Comparing Spark on FPGAs vs GPUs
Choose Your Weapon: Comparing Spark on FPGAs vs GPUsChoose Your Weapon: Comparing Spark on FPGAs vs GPUs
Choose Your Weapon: Comparing Spark on FPGAs vs GPUs
 
AWS Summit Bogotá Track Avanzado: EC2 avanzado
AWS Summit Bogotá Track Avanzado: EC2 avanzadoAWS Summit Bogotá Track Avanzado: EC2 avanzado
AWS Summit Bogotá Track Avanzado: EC2 avanzado
 
Making the most out of Heterogeneous Chips with CPU, GPU and FPGA
Making the most out of Heterogeneous Chips with CPU, GPU and FPGAMaking the most out of Heterogeneous Chips with CPU, GPU and FPGA
Making the most out of Heterogeneous Chips with CPU, GPU and FPGA
 
Programming Models for Heterogeneous Chips
Programming Models for  Heterogeneous ChipsProgramming Models for  Heterogeneous Chips
Programming Models for Heterogeneous Chips
 
Tackling Network Bottlenecks with Hardware Accelerations: Cloud vs. On-Premise
Tackling Network Bottlenecks with Hardware Accelerations: Cloud vs. On-PremiseTackling Network Bottlenecks with Hardware Accelerations: Cloud vs. On-Premise
Tackling Network Bottlenecks with Hardware Accelerations: Cloud vs. On-Premise
 
”Bare-Metal Container" presented at HPCC2016
”Bare-Metal Container" presented at HPCC2016”Bare-Metal Container" presented at HPCC2016
”Bare-Metal Container" presented at HPCC2016
 
AMD K6
AMD K6AMD K6
AMD K6
 
Using GPUs to handle Big Data with Java by Adam Roberts.
Using GPUs to handle Big Data with Java by Adam Roberts.Using GPUs to handle Big Data with Java by Adam Roberts.
Using GPUs to handle Big Data with Java by Adam Roberts.
 
Arm Neoverse market update_05122020.pdf
Arm Neoverse market update_05122020.pdfArm Neoverse market update_05122020.pdf
Arm Neoverse market update_05122020.pdf
 
The Rise of Parallel Computing
The Rise of Parallel ComputingThe Rise of Parallel Computing
The Rise of Parallel Computing
 
Pgopencl
PgopenclPgopencl
Pgopencl
 
PostgreSQL with OpenCL
PostgreSQL with OpenCLPostgreSQL with OpenCL
PostgreSQL with OpenCL
 
QEMU and Raspberry Pi. Instant Embedded Development
QEMU and Raspberry Pi. Instant Embedded DevelopmentQEMU and Raspberry Pi. Instant Embedded Development
QEMU and Raspberry Pi. Instant Embedded Development
 
Deep Dive on Delivering Amazon EC2 Instance Performance
Deep Dive on Delivering Amazon EC2 Instance PerformanceDeep Dive on Delivering Amazon EC2 Instance Performance
Deep Dive on Delivering Amazon EC2 Instance Performance
 
Computação acelerada – a era das ap us roberto brandão, ciência
Computação acelerada – a era das ap us   roberto brandão,  ciênciaComputação acelerada – a era das ap us   roberto brandão,  ciência
Computação acelerada – a era das ap us roberto brandão, ciência
 
Achieving the Ultimate Performance with KVM
Achieving the Ultimate Performance with KVMAchieving the Ultimate Performance with KVM
Achieving the Ultimate Performance with KVM
 

More from Igor Sfiligoi

Preparing Fusion codes for Perlmutter - CGYRO
Preparing Fusion codes for Perlmutter - CGYROPreparing Fusion codes for Perlmutter - CGYRO
Preparing Fusion codes for Perlmutter - CGYROIgor Sfiligoi
 
Comparing single-node and multi-node performance of an important fusion HPC c...
Comparing single-node and multi-node performance of an important fusion HPC c...Comparing single-node and multi-node performance of an important fusion HPC c...
Comparing single-node and multi-node performance of an important fusion HPC c...Igor Sfiligoi
 
The anachronism of whole-GPU accounting
The anachronism of whole-GPU accountingThe anachronism of whole-GPU accounting
The anachronism of whole-GPU accountingIgor Sfiligoi
 
Auto-scaling HTCondor pools using Kubernetes compute resources
Auto-scaling HTCondor pools using Kubernetes compute resourcesAuto-scaling HTCondor pools using Kubernetes compute resources
Auto-scaling HTCondor pools using Kubernetes compute resourcesIgor Sfiligoi
 
Speeding up bowtie2 by improving cache-hit rate
Speeding up bowtie2 by improving cache-hit rateSpeeding up bowtie2 by improving cache-hit rate
Speeding up bowtie2 by improving cache-hit rateIgor Sfiligoi
 
Performance Optimization of CGYRO for Multiscale Turbulence Simulations
Performance Optimization of CGYRO for Multiscale Turbulence SimulationsPerformance Optimization of CGYRO for Multiscale Turbulence Simulations
Performance Optimization of CGYRO for Multiscale Turbulence SimulationsIgor Sfiligoi
 
Comparing GPU effectiveness for Unifrac distance compute
Comparing GPU effectiveness for Unifrac distance computeComparing GPU effectiveness for Unifrac distance compute
Comparing GPU effectiveness for Unifrac distance computeIgor Sfiligoi
 
Managing Cloud networking costs for data-intensive applications by provisioni...
Managing Cloud networking costs for data-intensive applications by provisioni...Managing Cloud networking costs for data-intensive applications by provisioni...
Managing Cloud networking costs for data-intensive applications by provisioni...Igor Sfiligoi
 
Accelerating Key Bioinformatics Tasks 100-fold by Improving Memory Access
Accelerating Key Bioinformatics Tasks 100-fold by Improving Memory AccessAccelerating Key Bioinformatics Tasks 100-fold by Improving Memory Access
Accelerating Key Bioinformatics Tasks 100-fold by Improving Memory AccessIgor Sfiligoi
 
Using A100 MIG to Scale Astronomy Scientific Output
Using A100 MIG to Scale Astronomy Scientific OutputUsing A100 MIG to Scale Astronomy Scientific Output
Using A100 MIG to Scale Astronomy Scientific OutputIgor Sfiligoi
 
Using commercial Clouds to process IceCube jobs
Using commercial Clouds to process IceCube jobsUsing commercial Clouds to process IceCube jobs
Using commercial Clouds to process IceCube jobsIgor Sfiligoi
 
Modest scale HPC on Azure using CGYRO
Modest scale HPC on Azure using CGYROModest scale HPC on Azure using CGYRO
Modest scale HPC on Azure using CGYROIgor Sfiligoi
 
Data-intensive IceCube Cloud Burst
Data-intensive IceCube Cloud BurstData-intensive IceCube Cloud Burst
Data-intensive IceCube Cloud BurstIgor Sfiligoi
 
Scheduling a Kubernetes Federation with Admiralty
Scheduling a Kubernetes Federation with AdmiraltyScheduling a Kubernetes Federation with Admiralty
Scheduling a Kubernetes Federation with AdmiraltyIgor Sfiligoi
 
Accelerating microbiome research with OpenACC
Accelerating microbiome research with OpenACCAccelerating microbiome research with OpenACC
Accelerating microbiome research with OpenACCIgor Sfiligoi
 
Demonstrating a Pre-Exascale, Cost-Effective Multi-Cloud Environment for Scie...
Demonstrating a Pre-Exascale, Cost-Effective Multi-Cloud Environment for Scie...Demonstrating a Pre-Exascale, Cost-Effective Multi-Cloud Environment for Scie...
Demonstrating a Pre-Exascale, Cost-Effective Multi-Cloud Environment for Scie...Igor Sfiligoi
 
Porting and optimizing UniFrac for GPUs
Porting and optimizing UniFrac for GPUsPorting and optimizing UniFrac for GPUs
Porting and optimizing UniFrac for GPUsIgor Sfiligoi
 
Demonstrating 100 Gbps in and out of the public Clouds
Demonstrating 100 Gbps in and out of the public CloudsDemonstrating 100 Gbps in and out of the public Clouds
Demonstrating 100 Gbps in and out of the public CloudsIgor Sfiligoi
 
TransAtlantic Networking using Cloud links
TransAtlantic Networking using Cloud linksTransAtlantic Networking using Cloud links
TransAtlantic Networking using Cloud linksIgor Sfiligoi
 
Bursting into the public Cloud - Sharing my experience doing it at large scal...
Bursting into the public Cloud - Sharing my experience doing it at large scal...Bursting into the public Cloud - Sharing my experience doing it at large scal...
Bursting into the public Cloud - Sharing my experience doing it at large scal...Igor Sfiligoi
 

More from Igor Sfiligoi (20)

Preparing Fusion codes for Perlmutter - CGYRO
Preparing Fusion codes for Perlmutter - CGYROPreparing Fusion codes for Perlmutter - CGYRO
Preparing Fusion codes for Perlmutter - CGYRO
 
Comparing single-node and multi-node performance of an important fusion HPC c...
Comparing single-node and multi-node performance of an important fusion HPC c...Comparing single-node and multi-node performance of an important fusion HPC c...
Comparing single-node and multi-node performance of an important fusion HPC c...
 
The anachronism of whole-GPU accounting
The anachronism of whole-GPU accountingThe anachronism of whole-GPU accounting
The anachronism of whole-GPU accounting
 
Auto-scaling HTCondor pools using Kubernetes compute resources
Auto-scaling HTCondor pools using Kubernetes compute resourcesAuto-scaling HTCondor pools using Kubernetes compute resources
Auto-scaling HTCondor pools using Kubernetes compute resources
 
Speeding up bowtie2 by improving cache-hit rate
Speeding up bowtie2 by improving cache-hit rateSpeeding up bowtie2 by improving cache-hit rate
Speeding up bowtie2 by improving cache-hit rate
 
Performance Optimization of CGYRO for Multiscale Turbulence Simulations
Performance Optimization of CGYRO for Multiscale Turbulence SimulationsPerformance Optimization of CGYRO for Multiscale Turbulence Simulations
Performance Optimization of CGYRO for Multiscale Turbulence Simulations
 
Comparing GPU effectiveness for Unifrac distance compute
Comparing GPU effectiveness for Unifrac distance computeComparing GPU effectiveness for Unifrac distance compute
Comparing GPU effectiveness for Unifrac distance compute
 
Managing Cloud networking costs for data-intensive applications by provisioni...
Managing Cloud networking costs for data-intensive applications by provisioni...Managing Cloud networking costs for data-intensive applications by provisioni...
Managing Cloud networking costs for data-intensive applications by provisioni...
 
Accelerating Key Bioinformatics Tasks 100-fold by Improving Memory Access
Accelerating Key Bioinformatics Tasks 100-fold by Improving Memory AccessAccelerating Key Bioinformatics Tasks 100-fold by Improving Memory Access
Accelerating Key Bioinformatics Tasks 100-fold by Improving Memory Access
 
Using A100 MIG to Scale Astronomy Scientific Output
Using A100 MIG to Scale Astronomy Scientific OutputUsing A100 MIG to Scale Astronomy Scientific Output
Using A100 MIG to Scale Astronomy Scientific Output
 
Using commercial Clouds to process IceCube jobs
Using commercial Clouds to process IceCube jobsUsing commercial Clouds to process IceCube jobs
Using commercial Clouds to process IceCube jobs
 
Modest scale HPC on Azure using CGYRO
Modest scale HPC on Azure using CGYROModest scale HPC on Azure using CGYRO
Modest scale HPC on Azure using CGYRO
 
Data-intensive IceCube Cloud Burst
Data-intensive IceCube Cloud BurstData-intensive IceCube Cloud Burst
Data-intensive IceCube Cloud Burst
 
Scheduling a Kubernetes Federation with Admiralty
Scheduling a Kubernetes Federation with AdmiraltyScheduling a Kubernetes Federation with Admiralty
Scheduling a Kubernetes Federation with Admiralty
 
Accelerating microbiome research with OpenACC
Accelerating microbiome research with OpenACCAccelerating microbiome research with OpenACC
Accelerating microbiome research with OpenACC
 
Demonstrating a Pre-Exascale, Cost-Effective Multi-Cloud Environment for Scie...
Demonstrating a Pre-Exascale, Cost-Effective Multi-Cloud Environment for Scie...Demonstrating a Pre-Exascale, Cost-Effective Multi-Cloud Environment for Scie...
Demonstrating a Pre-Exascale, Cost-Effective Multi-Cloud Environment for Scie...
 
Porting and optimizing UniFrac for GPUs
Porting and optimizing UniFrac for GPUsPorting and optimizing UniFrac for GPUs
Porting and optimizing UniFrac for GPUs
 
Demonstrating 100 Gbps in and out of the public Clouds
Demonstrating 100 Gbps in and out of the public CloudsDemonstrating 100 Gbps in and out of the public Clouds
Demonstrating 100 Gbps in and out of the public Clouds
 
TransAtlantic Networking using Cloud links
TransAtlantic Networking using Cloud linksTransAtlantic Networking using Cloud links
TransAtlantic Networking using Cloud links
 
Bursting into the public Cloud - Sharing my experience doing it at large scal...
Bursting into the public Cloud - Sharing my experience doing it at large scal...Bursting into the public Cloud - Sharing my experience doing it at large scal...
Bursting into the public Cloud - Sharing my experience doing it at large scal...
 

Recently uploaded

SIEMENS: RAPUNZEL – A Tale About Knowledge Graph
SIEMENS: RAPUNZEL – A Tale About Knowledge GraphSIEMENS: RAPUNZEL – A Tale About Knowledge Graph
SIEMENS: RAPUNZEL – A Tale About Knowledge GraphNeo4j
 
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhi
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | DelhiFULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhi
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhisoniya singh
 
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking MenDelhi Call girls
 
Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Scott Keck-Warren
 
Benefits Of Flutter Compared To Other Frameworks
Benefits Of Flutter Compared To Other FrameworksBenefits Of Flutter Compared To Other Frameworks
Benefits Of Flutter Compared To Other FrameworksSoftradix Technologies
 
08448380779 Call Girls In Civil Lines Women Seeking Men
08448380779 Call Girls In Civil Lines Women Seeking Men08448380779 Call Girls In Civil Lines Women Seeking Men
08448380779 Call Girls In Civil Lines Women Seeking MenDelhi Call girls
 
Understanding the Laravel MVC Architecture
Understanding the Laravel MVC ArchitectureUnderstanding the Laravel MVC Architecture
Understanding the Laravel MVC ArchitecturePixlogix Infotech
 
Human Factors of XR: Using Human Factors to Design XR Systems
Human Factors of XR: Using Human Factors to Design XR SystemsHuman Factors of XR: Using Human Factors to Design XR Systems
Human Factors of XR: Using Human Factors to Design XR SystemsMark Billinghurst
 
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 3652toLead Limited
 
Key Features Of Token Development (1).pptx
Key  Features Of Token  Development (1).pptxKey  Features Of Token  Development (1).pptx
Key Features Of Token Development (1).pptxLBM Solutions
 
The Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxThe Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxMalak Abu Hammad
 
My Hashitalk Indonesia April 2024 Presentation
My Hashitalk Indonesia April 2024 PresentationMy Hashitalk Indonesia April 2024 Presentation
My Hashitalk Indonesia April 2024 PresentationRidwan Fadjar
 
Presentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreterPresentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreternaman860154
 
Making_way_through_DLL_hollowing_inspite_of_CFG_by_Debjeet Banerjee.pptx
Making_way_through_DLL_hollowing_inspite_of_CFG_by_Debjeet Banerjee.pptxMaking_way_through_DLL_hollowing_inspite_of_CFG_by_Debjeet Banerjee.pptx
Making_way_through_DLL_hollowing_inspite_of_CFG_by_Debjeet Banerjee.pptxnull - The Open Security Community
 
Next-generation AAM aircraft unveiled by Supernal, S-A2
Next-generation AAM aircraft unveiled by Supernal, S-A2Next-generation AAM aircraft unveiled by Supernal, S-A2
Next-generation AAM aircraft unveiled by Supernal, S-A2Hyundai Motor Group
 
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersEnhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersThousandEyes
 
How to convert PDF to text with Nanonets
How to convert PDF to text with NanonetsHow to convert PDF to text with Nanonets
How to convert PDF to text with Nanonetsnaman860154
 
Snow Chain-Integrated Tire for a Safe Drive on Winter Roads
Snow Chain-Integrated Tire for a Safe Drive on Winter RoadsSnow Chain-Integrated Tire for a Safe Drive on Winter Roads
Snow Chain-Integrated Tire for a Safe Drive on Winter RoadsHyundai Motor Group
 
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...Neo4j
 
SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024Scott Keck-Warren
 

Recently uploaded (20)

SIEMENS: RAPUNZEL – A Tale About Knowledge Graph
SIEMENS: RAPUNZEL – A Tale About Knowledge GraphSIEMENS: RAPUNZEL – A Tale About Knowledge Graph
SIEMENS: RAPUNZEL – A Tale About Knowledge Graph
 
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhi
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | DelhiFULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhi
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhi
 
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
 
Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024
 
Benefits Of Flutter Compared To Other Frameworks
Benefits Of Flutter Compared To Other FrameworksBenefits Of Flutter Compared To Other Frameworks
Benefits Of Flutter Compared To Other Frameworks
 
08448380779 Call Girls In Civil Lines Women Seeking Men
08448380779 Call Girls In Civil Lines Women Seeking Men08448380779 Call Girls In Civil Lines Women Seeking Men
08448380779 Call Girls In Civil Lines Women Seeking Men
 
Understanding the Laravel MVC Architecture
Understanding the Laravel MVC ArchitectureUnderstanding the Laravel MVC Architecture
Understanding the Laravel MVC Architecture
 
Human Factors of XR: Using Human Factors to Design XR Systems
Human Factors of XR: Using Human Factors to Design XR SystemsHuman Factors of XR: Using Human Factors to Design XR Systems
Human Factors of XR: Using Human Factors to Design XR Systems
 
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365
 
Key Features Of Token Development (1).pptx
Key  Features Of Token  Development (1).pptxKey  Features Of Token  Development (1).pptx
Key Features Of Token Development (1).pptx
 
The Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxThe Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptx
 
My Hashitalk Indonesia April 2024 Presentation
My Hashitalk Indonesia April 2024 PresentationMy Hashitalk Indonesia April 2024 Presentation
My Hashitalk Indonesia April 2024 Presentation
 
Presentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreterPresentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreter
 
Making_way_through_DLL_hollowing_inspite_of_CFG_by_Debjeet Banerjee.pptx
Making_way_through_DLL_hollowing_inspite_of_CFG_by_Debjeet Banerjee.pptxMaking_way_through_DLL_hollowing_inspite_of_CFG_by_Debjeet Banerjee.pptx
Making_way_through_DLL_hollowing_inspite_of_CFG_by_Debjeet Banerjee.pptx
 
Next-generation AAM aircraft unveiled by Supernal, S-A2
Next-generation AAM aircraft unveiled by Supernal, S-A2Next-generation AAM aircraft unveiled by Supernal, S-A2
Next-generation AAM aircraft unveiled by Supernal, S-A2
 
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersEnhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
 
How to convert PDF to text with Nanonets
How to convert PDF to text with NanonetsHow to convert PDF to text with Nanonets
How to convert PDF to text with Nanonets
 
Snow Chain-Integrated Tire for a Safe Drive on Winter Roads
Snow Chain-Integrated Tire for a Safe Drive on Winter RoadsSnow Chain-Integrated Tire for a Safe Drive on Winter Roads
Snow Chain-Integrated Tire for a Safe Drive on Winter Roads
 
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
 
SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024
 

O&C Meeting - Evaluation of ARM CPUs for IceCube available through Google Kubernetes Engine

  • 1. Evaluation of ARM CPUs for IceCube available through Google Kubernetes Engine Igor Sfiligoi – UCSD in collaboration with David Schultz & Benedikt Riedel - University of Wisconsin Madison CMS O&C Meeting, 9/28/22
  • 2. Benchmark setup • Google offered some credits to IceCube to be used on the Google Kubernetes Engine (GKE) • A managed Kubernetes setup on the Google Compute Cloud (GCP) • Google offers several generations of CPUs • x86 from both Intel and AMD, in both Hyper-threaded and not mode • ARM CPUs from Ampere Computing (no hyper-threading) • The main CPU-heavy application is CORSIKA • Cosmic air shower simulator • x86 binaries available on CVMFS • ARM binaries compiled in GKE CMS O&C Meeting, 9/28/22
  • 3. CPUs used • Being a Cloud platform, a lot of details are hidden • But this is what I can see: • t2a - one vCPU per core • Ampere Altra Arm Q64-30 processor with all-core frequency of 3.0 GHz • t2d - one vCPU per core • AMD EPYC 7B13 - 2.45 base 2.8 effective 3.5 max boost • c2d - two vCPU per core • AMD EPYC 7B13 - 2.45 base 2.8 effective 3.5 max boost • c2 - two vCPUs per core • Intel® Xeon® Gold 6253CL - 3.1 base 3.8 all core turbo 3.9 single core Same HW, different setup CMS O&C Meeting, 9/28/22
  • 4. Benchmark results ARM CPUs faster than all tested x86 CPUs Hyper-threading actually quite effective ARM CMS O&C Meeting, 9/28/22
  • 5. Cost effectiveness But AMD pricing definitely not in line with performance ARM CPUs more cost-effective than x86 CPUs ARM CMS O&C Meeting, 9/28/22
  • 6. Porting experience • x86 binaries already available, so a no brainer • We did not have ARM binaries, so we had to compile from source • Likely the first time it was done for most of the IceCube SW stack • The process was mostly painless • Some hardcoded x86-specific compiler parameters • Easy to replace with ARM-equivalents (but took a few passes to get them all) • CORSIKA itself built without a hitch • But we did need a bunch of helper binaries for the jobs to run • OS and the like came out of the box for both platforms • As standard docker images (we used k8s, remember?) CMS O&C Meeting, 9/28/22
  • 7. This is it for concrete data • The previous slides contain pretty much all that we have done • In the next few slides I will show a few reference data about power consumption I pulled from the Web • Take with a grain of salt! CMS O&C Meeting, 9/28/22
  • 8. CPU comparison • Ampere Altra Q64-30 • 64 cores at 3.0 GHz • 180W TDP • 8x 3200 DDR4 RAM • AMD EPYC 7B13 • CPU family 25, Stepping 1 • 60 cores, base freq 2.45 Ghz • 8x 3200 DDR4 RAM • Closest public TDP data is for EPYC 7713 at 225W TDP https://www.anandtech.com/show/16979/the-ampere- altra-max-review-pushing-it-to-128-cores-per-socket https://www.amd.com/en/products/cpu/amd-epyc-7713 https://cloud.google.com/compute/docs/general-purpose- machines#t2d_machines Intel Xeon 6253CL supposedly has 18 cores at 200W TDP CMS O&C Meeting, 9/28/22
  • 9. CPU comparison • Ampere Altra Q64-30 • 64 cores at 3.0 GHz • 180W TDP • 8x 3200 DDR4 RAM • AMD EPYC 7B13 • CPU family 25, Stepping 1 • 60 cores, base freq 2.45 Ghz • 8x 3200 DDR4 RAM • Closest public TDP data is for EPYC 7713 at 225W TDP https://www.anandtech.com/show/16979/the-ampere- altra-max-review-pushing-it-to-128-cores-per-socket https://www.amd.com/en/products/cpu/amd-epyc-7713 https://cloud.google.com/compute/docs/general-purpose- machines#t2d_machines Intel Xeon 6253CL supposedly has 18 cores at 200W TDP CPU power consumption is just a fraction of the total node power usage, of course! CMS O&C Meeting, 9/28/22
  • 10. CPU power effectiveness AMD EPYC in HT mode likely beats ARM CPU in power effectiveness. But ARM CPU more power efficient when non-HT mode is desired. Hyper-threading great for power effectiveness Disclaimer: Power usage based on estimated TDP CMS O&C Meeting, 9/28/22
  • 11. In summary • ARM CPUs now easily available in the commercial Clouds • Porting code to ARM not hard if you have the source code • In GKE, using IceCube CORSIKA simulation • ARM cores about as fast as the x86 cores (even a bit faster) • Priced cheaper, too • No hard data on power usage, but extrapolated data imply • ARM CPUs only modestly more power efficient that x86 competition CMS O&C Meeting, 9/28/22
  • 12. Acknowledgement • This work was partially funded by the U.S. National Science Foundation (NSF) under grants OAC-1826967, OAC-1541349, OPP- 2042807, CNS-1925001 and OAC- 2103963. • All Google Kubernetes Engine costs have been covered by Google- issued credits. CMS O&C Meeting, 9/28/22