AMD EPYC 9754 Bergamo SMT On/Off Comparison

Benchmarks by Michael Larabel for a future article (post 19th) looking at SMT on/off comparison toggled via BIOS. SMT comparison testing of AMD EPYC 9754 128-Core CPUs on Titanite with Ubuntu 22.04 LTS.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2307190-NE-BERGAMOSM27
Jump To Table - Results

View

Do Not Show Noisy Results
Do Not Show Results With Incomplete Data
Do Not Show Results With Little Change/Spread
List Notable Results

Limit displaying results to tests within:

C++ Boost Tests 2 Tests
Timed Code Compilation 5 Tests
C/C++ Compiler Tests 8 Tests
CPU Massive 14 Tests
Creator Workloads 11 Tests
Cryptography 4 Tests
Database Test Suite 2 Tests
Fortran Tests 5 Tests
Game Development 5 Tests
HPC - High Performance Computing 11 Tests
Common Kernel Benchmarks 2 Tests
Machine Learning 3 Tests
Molecular Dynamics 4 Tests
MPI Benchmarks 2 Tests
Multi-Core 23 Tests
NVIDIA GPU Compute 2 Tests
Intel oneAPI 6 Tests
OpenMPI Tests 8 Tests
Programmer / Developer System Benchmarks 5 Tests
Python Tests 8 Tests
Raytracing 2 Tests
Renderers 5 Tests
Scientific Computing 4 Tests
Software Defined Radio 2 Tests
Server 3 Tests
Server CPU Tests 12 Tests

Statistics

Show Overall Harmonic Mean(s)
Show Overall Geometric Mean
Show Geometric Means Per-Suite/Category
Show Wins / Losses Counts (Pie Chart)
Normalize Results
Remove Outliers Before Calculating Averages

Graph Settings

Force Line Graphs Where Applicable
Convert To Scalar Where Applicable
Disable Color Branding
Prefer Vertical Bar Graphs
No Box Plots
On Line Graphs With Missing Data, Connect The Line Gaps

Multi-Way Comparison

Condense Comparison
Transpose Comparison

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Hide
Result
Result
Identifier
View Logs
Performance Per
Dollar
Date
Run
  Test
  Duration
EPYC 9754 1P: SMT On
July 14 2023
  16 Hours, 51 Minutes
EPYC 9754 1P: SMT Off
July 13 2023
  15 Hours, 43 Minutes
EPYC 9754 2P: SMT On
July 11 2023
  14 Hours, 12 Minutes
EPYC 9754 2P: SMT Off
July 12 2023
  13 Hours, 40 Minutes
Invert Hiding All Results Option
  15 Hours, 6 Minutes

Only show results where is faster than
Only show results matching title/arguments (delimit multiple options with a comma):
Do not show results matching title/arguments (delimit multiple options with a comma):

EPYC 9754 1P: SMT On

Processor: AMD EPYC 9754 128-Core @ 2.25GHz (128 Cores / 256 Threads), Motherboard: AMD Titanite_4G (RTI1007B BIOS), Chipset: AMD Device 14a4, Memory: 768GB, Disk: 2 x 1920GB SAMSUNG MZWLJ1T9HBJR-00007, Graphics: ASPEED, Network: Broadcom NetXtreme BCM5720 PCIe

OS: Ubuntu 22.04, Kernel: 5.19.0-41-generic (x86_64), Desktop: GNOME Shell 42.5, Display Server: X Server 1.21.1.4, Vulkan: 1.3.224, Compiler: GCC 11.3.0, File-System: ext4, Screen Resolution: 1024x768

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-xKiWfi/gcc-11-11.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-xKiWfi/gcc-11-11.3.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v
Processor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xaa0010b
Python Notes: Python 3.10.6
Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

EPYC 9754 1P: SMT Off

Changed Processor to AMD EPYC 9754 128-Core @ 2.25GHz (128 Cores).

Security Change: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: disabled RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

EPYC 9754 2P: SMT On

Changed Processor to 2 x AMD EPYC 9754 128-Core @ 2.25GHz (256 Cores / 512 Threads).

Changed Memory to 1520GB.

EPYC 9754 2P: SMT Off

Changed Processor to 2 x AMD EPYC 9754 128-Core @ 2.25GHz (256 Cores).

Security Change: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: disabled RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

ProcessorMotherboardChipsetMemoryDiskGraphicsNetworkOSKernelDesktopDisplay ServerVulkanCompilerFile-SystemScreen ResolutionEPYC 9754 1PEPYC 9754 2P SMT On SMT Off SMT On SMT OffAMD EPYC 9754 128-Core @ 2.25GHz (128 Cores / 256 Threads)AMD Titanite_4G (RTI1007B BIOS)AMD Device 14a4768GB2 x 1920GB SAMSUNG MZWLJ1T9HBJR-00007ASPEEDBroadcom NetXtreme BCM5720 PCIeUbuntu 22.045.19.0-41-generic (x86_64)GNOME Shell 42.5X Server 1.21.1.41.3.224GCC 11.3.0ext41024x768AMD EPYC 9754 128-Core @ 2.25GHz (128 Cores)2 x AMD EPYC 9754 128-Core @ 2.25GHz (256 Cores / 512 Threads)1520GB2 x AMD EPYC 9754 128-Core @ 2.25GHz (256 Cores)OpenBenchmarking.orgKernel Details- Transparent Huge Pages: madviseCompiler Details- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-xKiWfi/gcc-11-11.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-xKiWfi/gcc-11-11.3.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v Processor Details- Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xaa0010bPython Details- Python 3.10.6Security Details- EPYC 9754 1P: SMT On: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 9754 1P: SMT Off: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: disabled RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 9754 2P: SMT On: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 9754 2P: SMT Off: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: disabled RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

openvino: Vehicle Detection FP16 - CPUopenssl: SHA256toybrot: TBBspecfem3d: Water-layered Halfspacecompress-7zip: Decompression Ratingdeepsparse: CV Detection, YOLOv5s COCO - Asynchronous Multi-Streamjohn-the-ripper: Blowfishjohn-the-ripper: bcryptdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Streamdeepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Streamdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Streamdeepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Streamembree: Pathtracer ISPC - Crowndeepsparse: NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Streamgraph500: 26openssl: ChaCha20embree: Pathtracer ISPC - Asian Dragonspecfem3d: Mount St. Helensgraph500: 26blender: Classroom - CPU-Onlynpb: LU.Cliquid-dsp: 512 - 256 - 512ospray-studio: 3 - 4K - 32 - Path Tracerospray-studio: 3 - 4K - 16 - Path Tracerospray-studio: 1 - 4K - 16 - Path Tracercp2k: H2O-DFT-LSospray-studio: 1 - 4K - 1 - Path Tracerospray-studio: 2 - 4K - 16 - Path Tracercloverleaf: Lagrangian-Eulerian Hydrodynamicsospray-studio: 3 - 4K - 1 - Path Tracerospray-studio: 2 - 4K - 32 - Path Traceropenssl: ChaCha20-Poly1305ospray-studio: 2 - 4K - 1 - Path Tracergraph500: 26ospray-studio: 1 - 4K - 32 - Path Tracerblender: Pabellon Barcelona - CPU-Onlyjohn-the-ripper: WPA PSKxmrig: Wownero - 1Mastcenc: Exhaustiveopenvino: Person Detection FP32 - CPUopenvino: Machine Translation EN To DE FP16 - CPUopenvino: Person Detection FP16 - CPUspecfem3d: Layered Halfspaceblender: Barbershop - CPU-Onlyblender: BMW27 - CPU-Onlyhelsing: 14 digitgraph500: 26openssl: RSA4096npb: MG.Castcenc: Fastospray: gravity_spheres_volume/dim_512/ao/real_timeopenssl: RSA4096john-the-ripper: MD5ospray: particle_volume/scivis/real_timeblender: Fishy Cat - CPU-Onlyospray: particle_volume/ao/real_timeospray: gravity_spheres_volume/dim_512/scivis/real_timeopenssl: SHA512openvino: Face Detection FP16 - CPUopenssl: AES-128-GCMopenssl: AES-256-GCMopenvino: Machine Translation EN To DE FP16 - CPUopenvino: Face Detection FP16-INT8 - CPUopenvino: Face Detection FP16-INT8 - CPUopenvino: Face Detection FP16 - CPUdeepsparse: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Asynchronous Multi-Streamastcenc: Thoroughopenvino: Weld Porosity Detection FP16-INT8 - CPUopenvino: Weld Porosity Detection FP16 - CPUnamd: ATPase Simulation - 327,506 Atomsliquid-dsp: 256 - 256 - 512openvino: Person Vehicle Bike Detection FP16 - CPUbuild-linux-kernel: allmodconfigopenvino: Weld Porosity Detection FP16 - CPUprimesieve: 1e13libxsmm: 256toybrot: OpenMPopenvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPUminibude: OpenMP - BM2minibude: OpenMP - BM2npb: IS.Dnpb: BT.Cheffte: r2c - FFTW - float - 512npb: SP.Cluxcorerender: LuxCore Benchmark - CPUheffte: c2c - FFTW - float - 512primesieve: 1e12openvino: Person Vehicle Bike Detection FP16 - CPUtensorflow: CPU - 256 - GoogLeNetappleseed: Material Testernpb: FT.Ccompress-7zip: Compression Ratingopenvkl: vklBenchmark ISPCtensorflow: CPU - 512 - ResNet-50openvino: Person Detection FP16 - CPUopenvino: Person Detection FP32 - CPUtensorflow: CPU - 512 - GoogLeNetopenvino: Age Gender Recognition Retail 0013 FP16 - CPUnpb: CG.Cbuild-linux-kernel: defconfigluxcorerender: DLSC - CPUdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Streamdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Streamdeepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Streamdeepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Streamdeepsparse: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Asynchronous Multi-Streamoidn: RTLightmap.hdr.4096x4096 - CPU-Onlydeepsparse: NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Streamdeepsparse: CV Detection, YOLOv5s COCO - Asynchronous Multi-Streammysqlslap: 2048appleseed: Emilyoidn: RT.ldr_alb_nrm.3840x2160 - CPU-Onlyopenvino: Age Gender Recognition Retail 0013 FP16 - CPUaircrack-ng: oidn: RT.hdr_alb_nrm.3840x2160 - CPU-Onlytensorflow: CPU - 256 - AlexNetmysqlslap: 4096build-llvm: Ninjatensorflow: CPU - 512 - AlexNetbuild-nodejs: Time To Compiletensorflow: CPU - 256 - ResNet-50minife: Smallopenvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPUappleseed: Disney Materialbuild-gem5: Time To Compilebuild-llvm: Unix Makefilesbuild-godot: Time To Compileopenvino: Weld Porosity Detection FP16-INT8 - CPUnekrs: Kershawopenvino: Vehicle Detection FP16-INT8 - CPUopenvino: Vehicle Detection FP16-INT8 - CPUopenvino: Vehicle Detection FP16 - CPUdeepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Streamdeepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Streamdeepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Streamdeepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Streampgbench: 1000 - 800 - Read Only - Average Latencypgbench: 1000 - 800 - Read Onlystockfish: Total Timeluxcorerender: Rainbow Colors and Prism - CPUluxcorerender: Orange Juice - CPUsrsran: PUSCH Processor Benchmark, Throughput Totalxmrig: Monero - 1Mnekrs: TurboPipe Periodicspecfem3d: Homogeneous Halfspacespecfem3d: Tomographic Modelheffte: r2c - FFTW - double - 512heffte: c2c - FFTW - double - 512libxsmm: 128minibude: OpenMP - BM1minibude: OpenMP - BM1npb: SP.Bnpb: EP.DEPYC 9754 1PEPYC 9754 2P SMT On SMT Off SMT On SMT Off43.73163633625553359117.194492779791787417.118621611521606473.2077968.626773.3807624.4789125.5813316.0466445912000659346857987157.65046.21637984988024900031.12279662.5517837000003297216484137595012.268611396312.001032278854624158373208733334450002753839.2781037574803.68.17432378.07582.082377.7915.964931688116.5412.7750.4738578900001890935.3128129.561190.754932.675054195.12031266730.816516.4930.850731.9245530058793301049.3111696737355571012537766210110.00541.29117.8860.771376.214275.083111794.326067.780.2070216967666676148.89227.51710.5321.2863331.7408185400.885972.643238.9055300.29292243.61245.542131909.9112.18128.1241.94410.40504.09166.676131140791.527262711396122.7726.6126.57416.03120515.1145686.8826.22516.34859.8653859.87765.9747102.226546.43181.74201.6258152.9895780122.5877093.620.83171120.3543.621422.08655125.3641628.80113.704118.4551784.11.2144.30492161.648211.079105.79710.84580863666712.075311.481464.29499.5866126.9310267.7088240.86670.95285556936503434920.8824.478389.024409.425384069239.4043485007.48067470666.341834.84512713.4237.7635944.062149355.5413264.7911.49111414557280559013.419454501510470504.357816326316322097.09131275.770796.9697812.102385.2893404.2338493535000550700307433107.36085.03003158092832000038.44289518.1414147000004337121666180044957.6861127182699.2713583653939278268998711463637500003592650.0167621863182.27.31471102.82545.131105.9615.028121929147.2215.1557.9568936240001799293.0136942.131278.646625.717856647.31675166723.721020.4723.733325.512151804333637513.23115283892806399703462188358.71268.82119.0162.181780.442168.307211710.595837.620.2059513139333335118.29177.7585.4721.1323813.4624271673.855903.205236.1285315.15298801.44248.557133415.428.88128.5971.7966.24525.39167.900597147448.725927411107123.8828.7428.83429.16113162.2948672.2422.98413.27644.9768645.001949.511177.814835.55491.72156.0563125.4467783123.2983963.570.70149056.9533.571375.33695119.5761526.81116.122121.4751741.01.3038.492263148.484213.907102.58810.9157342666674.746744.662784.67468.7809134.5717259.2779246.37320.91787772227272294014.3720.9820430.951218.925860833336.0927502994.93726550667.990735.40492696.5234.6845867.108161475.2414274.5311.0332792603851320149.7783814371353957797.6248409850407860139.88211868.3310139.65901190.7168210.0714602.26029604710001317549954027255.98643.906675095172403000016.28591505.17333016666718455925376982143.513482781721.65582157319098173601104956725410001561921.751524533142082.715.93051552.151159.991559.959.99744757069.037.1227.28015711000003782091.8249109.09610.113753.8295108490.53487933349.23119.8749.132852.7318106049655203526.982339890700107201931707032755.12270.93235.65121.032627.3142134.885322954.4411299.320.1064625444000009889.61145.9295.6411.1206112.63321133931.537888.085315.5249849.01491231.83433.601224243.289.86221.7651.5086.46329.22211432.899258201720172.3640.6740.90538.52168372.8967554.7420.34418.61902.8593902.779268.3638107.183548.62062.35211.4848159.8861580164.5671644.780.624.761225.41545107.7211770.9193.271124.9862774.21.08152.586199.365100.79011.004.8313214.205810.41521.3108242.9589229.1624556.94110.97482787858238692419.0234.4517891.486533.74.7278301483.782741798207.197109.6454976.7253.1236328.069236490.7623705.4010.2622226942886729766.2180692579130911058.6776320885317046182.59742409.8035182.44531541.9512146.2051770.139210759000001100453394570178.30462.630122599207888000020.24658754.2126109333332423812127100462363.5796311023215.15763205457844895705236457701800002013327.471301500100754.314.40991579.141174.991545.237.47077293085.548.4550.05418151600003598946.3268721.05693.396144.7104113251.43022166741.553111.7241.656744.2889102232233590526.462307524535457199823414886354.42271.24235.37121.082730.4980127.208822955.6411373.950.1396925426333339878.88118.0995.6111.1526373.03671118225.5510989.938439.5978635.30536518.74430.265231041.576.97223.5841.1606.47452.99265.885808224178.107714621530189.4041.0540.20634.13142135.5866822.9718.47414.45676.8128676.434251.810381.199445.87652.09162.0191118.1285591159.9066144.310.67128050.2194.351581.6657999.2151908.7693.952146.7453798.61.1340.570691148.376198.750100.24011.104.8613141.516242.68426.7362290.7562212.6821590.87891.01878596844702314313.5125.1136573.885946.03.4518301692.709205428210.933112.4134505.5439.06710976.682246272.7925983.74OpenBenchmarking.org

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Vehicle Detection FP16 - Device: CPUSMT OnSMT Off1020304050SE +/- 0.55, N = 14SE +/- 0.10, N = 13SE +/- 0.15, N = 15SE +/- 0.12, N = 1343.7311.4911.0310.261. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2022.3CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 10.52 / Avg: 252.92 / Max: 325.99Min: 20.5 / Avg: 276.2 / Max: 333.13Min: 42.88 / Avg: 597.65 / Max: 674.79Min: 193.73 / Avg: 615.75 / Max: 674.39
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Vehicle Detection FP16 - Device: CPUSMT OnSMT Off918273645Min: 36.74 / Avg: 43.73 / Max: 45.01Min: 10.35 / Avg: 11.49 / Max: 11.73Min: 9 / Avg: 11.03 / Max: 11.38Min: 8.78 / Avg: 10.26 / Max: 10.541. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: SHA256SMT OnSMT Off70000M140000M210000M280000M350000MSE +/- 161352968.33, N = 3SE +/- 15184699.42, N = 3SE +/- 207719324.78, N = 3SE +/- 224959781.68, N = 31636336255531114145572803279260385132222694288671. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgbyte/s Per Watt, More Is BetterOpenSSL 3.1Algorithm: SHA256SMT OnSMT Off110M220M330M440M550M491268818.21390668971.64485100025.27368054429.03
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenSSL 3.1CPU Power Consumption MonitorSMT OnSMT Off130260390520650Min: 20.72 / Avg: 333.08 / Max: 356.22Min: 20.37 / Avg: 285.19 / Max: 341.28Min: 41.96 / Avg: 676 / Max: 716.17Min: 42.55 / Avg: 603.9 / Max: 697.6
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: SHA256SMT OnSMT Off60000M120000M180000M240000M300000MMin: 163393399580 / Avg: 163633625553.33 / Max: 163940347630Min: 111390044420 / Avg: 111414557280 / Max: 111442339980Min: 327516173060 / Avg: 327926038513.33 / Max: 328189705270Min: 221819514970 / Avg: 222269428866.67 / Max: 2224963414001. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

toyBrot Fractal Generator

ToyBrot is a Mandelbrot fractal generator supporting C++ threads/tasks, OpenMP, Intel Threaded Building Blocks (TBB), and other targets. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BettertoyBrot Fractal Generator 2020-11-18Implementation: TBBSMT OnSMT Off12002400360048006000SE +/- 21.75, N = 9SE +/- 43.30, N = 15SE +/- 23.04, N = 15SE +/- 31.80, N = 1535915590201429761. (CXX) g++ options: -O3 -lpthread -lm -lgcc -lgcc_s -lc
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BettertoyBrot Fractal Generator 2020-11-18CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 20.91 / Avg: 150.09 / Max: 303.56Min: 20.03 / Avg: 161.83 / Max: 283.89Min: 41.02 / Avg: 232.68 / Max: 605.7Min: 40.87 / Avg: 257.24 / Max: 584.5
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BettertoyBrot Fractal Generator 2020-11-18Implementation: TBBSMT OnSMT Off10002000300040005000Min: 3519 / Avg: 3590.67 / Max: 3718Min: 5365 / Avg: 5589.53 / Max: 5923Min: 1863 / Avg: 2014.33 / Max: 2157Min: 2795 / Avg: 2975.6 / Max: 31591. (CXX) g++ options: -O3 -lpthread -lm -lgcc -lgcc_s -lc

SPECFEM3D

simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.0Model: Water-layered HalfspaceSMT OnSMT Off48121620SE +/- 0.142353904, N = 3SE +/- 0.021567327, N = 3SE +/- 0.063515594, N = 4SE +/- 0.062799204, N = 1517.19449277913.4194545019.7783814376.2180692571. (F9X) gfortran options: -O2 -fopenmp -std=f2003 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterSPECFEM3D 4.0CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.6 / Avg: 249.37 / Max: 338.57Min: 20.04 / Avg: 228.45 / Max: 329.73Min: 41.18 / Avg: 415.4 / Max: 678.2Min: 41.07 / Avg: 358.64 / Max: 670.26
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.0Model: Water-layered HalfspaceSMT OnSMT Off48121620Min: 16.98 / Avg: 17.19 / Max: 17.46Min: 13.38 / Avg: 13.42 / Max: 13.45Min: 9.66 / Avg: 9.78 / Max: 9.96Min: 6 / Avg: 6.22 / Max: 6.881. (F9X) gfortran options: -O2 -fopenmp -std=f2003 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

7-Zip Compression

This is a test of 7-Zip compression/decompression with its integrated benchmark feature. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 22.01Test: Decompression RatingSMT OnSMT Off300K600K900K1200K1500KSE +/- 1608.29, N = 3SE +/- 429.17, N = 3SE +/- 5499.45, N = 3SE +/- 1897.81, N = 379178751047013539579130911. (CXX) g++ options: -lpthread -ldl -O2 -fPIC
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgMIPS Per Watt, More Is Better7-Zip Compression 22.01Test: Decompression RatingSMT OnSMT Off60012001800240030002973.582063.782700.351933.66
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is Better7-Zip Compression 22.01CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.05 / Avg: 266.27 / Max: 350.56Min: 20.34 / Avg: 247.35 / Max: 330.87Min: 41.89 / Avg: 501.4 / Max: 701.41Min: 41.8 / Avg: 472.21 / Max: 667.59
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 22.01Test: Decompression RatingSMT OnSMT Off200K400K600K800K1000KMin: 789740 / Avg: 791786.67 / Max: 794959Min: 509612 / Avg: 510470.33 / Max: 510901Min: 1348033 / Avg: 1353957.33 / Max: 1364945Min: 910040 / Avg: 913091.33 / Max: 9165721. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-StreamSMT OnSMT Off2004006008001000SE +/- 0.46, N = 3SE +/- 5.07, N = 15SE +/- 0.32, N = 3SE +/- 0.90, N = 3417.12504.36797.621058.68
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-StreamSMT OnSMT Off2004006008001000Min: 416.27 / Avg: 417.12 / Max: 417.84Min: 493.02 / Avg: 504.36 / Max: 559.98Min: 797.19 / Avg: 797.62 / Max: 798.25Min: 1056.88 / Avg: 1058.68 / Max: 1059.73

John The Ripper

This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: BlowfishSMT OnSMT Off90K180K270K360K450KSE +/- 117.40, N = 3SE +/- 12.67, N = 3SE +/- 3726.56, N = 3SE +/- 1247.49, N = 32161151632634098503208851. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgReal C/S Per Watt, More Is BetterJohn The Ripper 2023.03.14Test: BlowfishSMT OnSMT Off170340510680850768.10632.58726.35582.13
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: BlowfishSMT OnSMT Off70K140K210K280K350KMin: 215884 / Avg: 216114.67 / Max: 216268Min: 163238 / Avg: 163263.33 / Max: 163276Min: 402400 / Avg: 409850 / Max: 413762Min: 318412 / Avg: 320885.33 / Max: 3224061. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: bcryptSMT OnSMT Off90K180K270K360K450KSE +/- 92.54, N = 3SE +/- 33.22, N = 3SE +/- 2298.36, N = 3SE +/- 1964.45, N = 32160641632204078603170461. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgReal C/S Per Watt, More Is BetterJohn The Ripper 2023.03.14Test: bcryptSMT OnSMT Off170340510680850770.87642.02736.92585.47
EPYC 9754 2PEPYC 9754 1POpenBenchmarking.orgWatts, Fewer Is BetterJohn The Ripper 2023.03.14CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 42.39 / Avg: 564.26 / Max: 642.19Min: 42.35 / Avg: 551.23 / Max: 631.68Min: 42.38 / Avg: 553.47 / Max: 644.02Min: 40.81 / Avg: 541.52 / Max: 627.33Min: 20.76 / Avg: 258.09 / Max: 293.57Min: 20.74 / Avg: 281.36 / Max: 322.03Min: 20.17 / Avg: 280.29 / Max: 320.86Min: 20 / Avg: 254.23 / Max: 291.94
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: bcryptSMT OnSMT Off70K140K210K280K350KMin: 215884 / Avg: 216063.67 / Max: 216192Min: 163161 / Avg: 163219.67 / Max: 163276Min: 405164 / Avg: 407859.67 / Max: 412432Min: 313651 / Avg: 317045.67 / Max: 3204561. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-StreamSMT OnSMT Off4080120160200SE +/- 0.25, N = 3SE +/- 0.06, N = 3SE +/- 0.08, N = 3SE +/- 0.18, N = 373.2197.09139.88182.60
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-StreamSMT OnSMT Off306090120150Min: 72.72 / Avg: 73.21 / Max: 73.55Min: 96.98 / Avg: 97.09 / Max: 97.17Min: 139.73 / Avg: 139.88 / Max: 139.99Min: 182.33 / Avg: 182.6 / Max: 182.95

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-StreamSMT OnSMT Off5001000150020002500SE +/- 0.74, N = 3SE +/- 5.02, N = 3SE +/- 1.60, N = 3SE +/- 1.63, N = 3968.631275.771868.332409.80
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-StreamSMT OnSMT Off400800120016002000Min: 967.57 / Avg: 968.63 / Max: 970.05Min: 1269.19 / Avg: 1275.77 / Max: 1285.63Min: 1865.16 / Avg: 1868.33 / Max: 1870.28Min: 2407.04 / Avg: 2409.8 / Max: 2412.69

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-StreamSMT OnSMT Off4080120160200SE +/- 0.25, N = 3SE +/- 0.09, N = 3SE +/- 0.15, N = 3SE +/- 0.07, N = 373.3896.97139.66182.45
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-StreamSMT OnSMT Off306090120150Min: 72.87 / Avg: 73.38 / Max: 73.67Min: 96.85 / Avg: 96.97 / Max: 97.15Min: 139.36 / Avg: 139.66 / Max: 139.85Min: 182.34 / Avg: 182.45 / Max: 182.58

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-StreamSMT OnSMT Off30060090012001500SE +/- 0.29, N = 3SE +/- 3.93, N = 3SE +/- 1.02, N = 3SE +/- 1.38, N = 3624.48812.101190.721541.95
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-StreamSMT OnSMT Off30060090012001500Min: 624.02 / Avg: 624.48 / Max: 625.02Min: 804.95 / Avg: 812.1 / Max: 818.52Min: 1188.75 / Avg: 1190.72 / Max: 1192.13Min: 1539.54 / Avg: 1541.95 / Max: 1544.32

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs (and GPUs via SYCL) and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.1Binary: Pathtracer ISPC - Model: CrownSMT OnSMT Off50100150200250SE +/- 0.13, N = 7SE +/- 0.05, N = 6SE +/- 0.16, N = 9SE +/- 0.13, N = 7125.5885.29210.07146.21
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterEmbree 4.1Binary: Pathtracer ISPC - Model: CrownSMT OnSMT Off0.15030.30060.45090.60120.75150.6550.4160.6680.420
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterEmbree 4.1CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.54 / Avg: 191.71 / Max: 332.64Min: 19.75 / Avg: 204.9 / Max: 312.76Min: 41.29 / Avg: 314.33 / Max: 666.86Min: 40.12 / Avg: 347.93 / Max: 636
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.1Binary: Pathtracer ISPC - Model: CrownSMT OnSMT Off4080120160200Min: 125.17 / Avg: 125.58 / Max: 126.22Min: 85.05 / Avg: 85.29 / Max: 85.42Min: 209.36 / Avg: 210.07 / Max: 210.97Min: 145.74 / Avg: 146.21 / Max: 146.72

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-StreamSMT OnSMT Off170340510680850SE +/- 1.02, N = 3SE +/- 2.15, N = 3SE +/- 0.72, N = 3SE +/- 0.92, N = 3316.05404.23602.26770.14
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-StreamSMT OnSMT Off140280420560700Min: 314.68 / Avg: 316.05 / Max: 318.04Min: 400.76 / Avg: 404.23 / Max: 408.16Min: 600.87 / Avg: 602.26 / Max: 603.3Min: 768.31 / Avg: 770.14 / Max: 771.07

Graph500

This is a benchmark of the reference implementation of Graph500, an HPC benchmark focused on data intensive loads and commonly tested on supercomputers for complex data problems. Graph500 primarily stresses the communication subsystem of the hardware under test. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgsssp max_TEPS, More Is BetterGraph500 3.0Scale: 26SMT OnSMT Off200M400M600M800M1000M44591200049353500096047100010759000001. (CC) gcc options: -fcommon -O3 -lpthread -lm -lmpi
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgsssp max_TEPS Per Watt, More Is BetterGraph500 3.0Scale: 26SMT OnSMT Off400K800K1200K1600K2000K1520544.671719832.821513281.141664973.76
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterGraph500 3.0CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.27 / Avg: 293.26 / Max: 336.48Min: 19.89 / Avg: 286.97 / Max: 330.07Min: 41.39 / Avg: 634.69 / Max: 685.66Min: 40.03 / Avg: 646.2 / Max: 688.13

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: ChaCha20SMT OnSMT Off300000M600000M900000M1200000M1500000MSE +/- 23916253.09, N = 3SE +/- 70893114.79, N = 3SE +/- 46014499.85, N = 3SE +/- 107897909.37, N = 3659346857987550700307433131754995402711004533945701. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgbyte/s Per Watt, More Is BetterOpenSSL 3.1Algorithm: ChaCha20SMT OnSMT Off400M800M1200M1600M2000M1812397903.611787936411.251804203926.061666389521.27
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenSSL 3.1CPU Power Consumption MonitorSMT OnSMT Off140280420560700Min: 21.23 / Avg: 363.8 / Max: 392.13Min: 20.75 / Avg: 308.01 / Max: 338.71Min: 44.02 / Avg: 730.27 / Max: 792.36Min: 42.08 / Avg: 660.38 / Max: 727.38
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: ChaCha20SMT OnSMT Off200000M400000M600000M800000M1000000MMin: 659303944470 / Avg: 659346857986.67 / Max: 659386611850Min: 550561616150 / Avg: 550700307433.33 / Max: 550795168700Min: 1317482271060 / Avg: 1317549954026.7 / Max: 1317637797960Min: 1100253662140 / Avg: 1100453394570 / Max: 11006240145401. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs (and GPUs via SYCL) and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.1Binary: Pathtracer ISPC - Model: Asian DragonSMT OnSMT Off60120180240300SE +/- 0.09, N = 8SE +/- 0.09, N = 6SE +/- 0.48, N = 9SE +/- 0.26, N = 8157.65107.36255.99178.30
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterEmbree 4.1Binary: Pathtracer ISPC - Model: Asian DragonSMT OnSMT Off0.20810.41620.62430.83241.04050.9250.5840.8950.570
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterEmbree 4.1CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 20.48 / Avg: 170.51 / Max: 326.15Min: 20.14 / Avg: 183.96 / Max: 309.98Min: 41.35 / Avg: 286.01 / Max: 649.79Min: 40.75 / Avg: 312.56 / Max: 630.49
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.1Binary: Pathtracer ISPC - Model: Asian DragonSMT OnSMT Off50100150200250Min: 157.31 / Avg: 157.65 / Max: 158.17Min: 107.1 / Avg: 107.36 / Max: 107.68Min: 253.76 / Avg: 255.99 / Max: 257.89Min: 177.2 / Avg: 178.3 / Max: 179.26

SPECFEM3D

simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.0Model: Mount St. HelensSMT OnSMT Off246810SE +/- 0.034771470, N = 5SE +/- 0.069356320, N = 12SE +/- 0.014293172, N = 5SE +/- 0.013893024, N = 56.2163798495.0300315803.9066750952.6301225991. (F9X) gfortran options: -O2 -fopenmp -std=f2003 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterSPECFEM3D 4.0CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.64 / Avg: 185.39 / Max: 336.6Min: 20.51 / Avg: 173.78 / Max: 324.55Min: 42.65 / Avg: 310.22 / Max: 679.55Min: 43.56 / Avg: 283.61 / Max: 651.99
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.0Model: Mount St. HelensSMT OnSMT Off246810Min: 6.11 / Avg: 6.22 / Max: 6.31Min: 4.62 / Avg: 5.03 / Max: 5.31Min: 3.87 / Avg: 3.91 / Max: 3.96Min: 2.61 / Avg: 2.63 / Max: 2.681. (F9X) gfortran options: -O2 -fopenmp -std=f2003 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

Graph500

This is a benchmark of the reference implementation of Graph500, an HPC benchmark focused on data intensive loads and commonly tested on supercomputers for complex data problems. Graph500 primarily stresses the communication subsystem of the hardware under test. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgbfs max_TEPS, More Is BetterGraph500 3.0Scale: 26SMT OnSMT Off400M800M1200M1600M2000M880249000928320000172403000020788800001. (CC) gcc options: -fcommon -O3 -lpthread -lm -lmpi

Blender

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.6Blend File: Classroom - Compute: CPU-OnlySMT OnSMT Off918273645SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.03, N = 3SE +/- 0.04, N = 331.1238.4416.2820.24
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterBlender 3.6CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.89 / Avg: 294.4 / Max: 340.14Min: 20.53 / Avg: 290.08 / Max: 328.71Min: 41.98 / Avg: 525.55 / Max: 684.07Min: 195.46 / Avg: 557.64 / Max: 654.85
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.6Blend File: Classroom - Compute: CPU-OnlySMT OnSMT Off816243240Min: 31.03 / Avg: 31.12 / Max: 31.16Min: 38.39 / Avg: 38.44 / Max: 38.52Min: 16.22 / Avg: 16.28 / Max: 16.34Min: 20.16 / Avg: 20.24 / Max: 20.3

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: LU.CSMT OnSMT Off140K280K420K560K700KSE +/- 2132.06, N = 6SE +/- 1485.96, N = 6SE +/- 7199.59, N = 15SE +/- 4916.03, N = 15279662.55289518.14591505.17658754.211. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s Per Watt, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: LU.CSMT OnSMT Off50010001500200025001448.631504.971821.352122.50
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNAS Parallel Benchmarks 3.4CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.31 / Avg: 193.05 / Max: 313.96Min: 19.72 / Avg: 192.37 / Max: 313.77Min: 40.56 / Avg: 324.76 / Max: 663.39Min: 40.61 / Avg: 310.37 / Max: 663.69
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: LU.CSMT OnSMT Off110K220K330K440K550KMin: 271434.32 / Avg: 279662.55 / Max: 286184.34Min: 284375.68 / Avg: 289518.14 / Max: 293094.64Min: 545882.78 / Avg: 591505.17 / Max: 641830.47Min: 625765.14 / Avg: 658754.21 / Max: 688059.031. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 512 - Buffer Length: 256 - Filter Length: 512SMT OnSMT Off700M1400M2100M2800M3500MSE +/- 1814754.35, N = 3SE +/- 2946183.97, N = 3SE +/- 569600.25, N = 3SE +/- 3555434.03, N = 317837000001414700000333016666726109333331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgsamples/s Per Watt, More Is BetterLiquid-DSP 1.6Threads: 512 - Buffer Length: 256 - Filter Length: 512SMT OnSMT Off1.4M2.8M4.2M5.6M7M6669884.785445501.586142943.895062810.80
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterLiquid-DSP 1.6CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 20.83 / Avg: 267.43 / Max: 313.4Min: 20.3 / Avg: 259.79 / Max: 301.25Min: 43.17 / Avg: 542.11 / Max: 636.19Min: 42.64 / Avg: 515.71 / Max: 606.26
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 512 - Buffer Length: 256 - Filter Length: 512SMT OnSMT Off600M1200M1800M2400M3000MMin: 1780900000 / Avg: 1783700000 / Max: 1787100000Min: 1408900000 / Avg: 1414700000 / Max: 1418500000Min: 3329500000 / Avg: 3330166666.67 / Max: 3331300000Min: 2604600000 / Avg: 2610933333.33 / Max: 26169000001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 3 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path TracerSMT OnSMT Off9K18K27K36K45KSE +/- 17.58, N = 3SE +/- 76.61, N = 3SE +/- 34.44, N = 3SE +/- 17.91, N = 3329724337118455242381. (CXX) g++ options: -O3 -lm -ldl
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOSPRay Studio 0.11CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.32 / Avg: 291.25 / Max: 338.09Min: 20.81 / Avg: 290.35 / Max: 325.63Min: 42.82 / Avg: 625.58 / Max: 686.92Min: 42.38 / Avg: 568.08 / Max: 654.97
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 3 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path TracerSMT OnSMT Off8K16K24K32K40KMin: 32945 / Avg: 32972 / Max: 33005Min: 43234 / Avg: 43370.67 / Max: 43499Min: 18388 / Avg: 18454.67 / Max: 18503Min: 24208 / Avg: 24238.33 / Max: 242701. (CXX) g++ options: -O3 -lm -ldl

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 3 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path TracerSMT OnSMT Off5K10K15K20K25KSE +/- 11.89, N = 3SE +/- 15.65, N = 3SE +/- 27.17, N = 3SE +/- 8.65, N = 316484216669253121271. (CXX) g++ options: -O3 -lm -ldl
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOSPRay Studio 0.11CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.56 / Avg: 282.67 / Max: 337.69Min: 20.73 / Avg: 274.66 / Max: 324.6Min: 44.72 / Avg: 616.86 / Max: 687.3Min: 41.61 / Avg: 566.27 / Max: 654
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 3 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path TracerSMT OnSMT Off4K8K12K16K20KMin: 16461 / Avg: 16484.33 / Max: 16500Min: 21638 / Avg: 21666.33 / Max: 21692Min: 9207 / Avg: 9252.67 / Max: 9301Min: 12115 / Avg: 12127.33 / Max: 121441. (CXX) g++ options: -O3 -lm -ldl

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 1 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path TracerSMT OnSMT Off4K8K12K16K20KSE +/- 5.78, N = 3SE +/- 18.26, N = 3SE +/- 7.80, N = 3SE +/- 14.52, N = 313759180047698100461. (CXX) g++ options: -O3 -lm -ldl
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOSPRay Studio 0.11CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.61 / Avg: 305.02 / Max: 338.57Min: 20.9 / Avg: 295.01 / Max: 325.94Min: 42.72 / Avg: 617.3 / Max: 686.55Min: 42.66 / Avg: 587.43 / Max: 654.45
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 1 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path TracerSMT OnSMT Off3K6K9K12K15KMin: 13751 / Avg: 13758.67 / Max: 13770Min: 17978 / Avg: 18003.67 / Max: 18039Min: 7684 / Avg: 7697.67 / Max: 7711Min: 10017 / Avg: 10045.67 / Max: 100641. (CXX) g++ options: -O3 -lm -ldl

CP2K Molecular Dynamics

CP2K is an open-source molecular dynamics software package focused on quantum chemistry and solid-state physics. More details on the CP2K benchmark test cases and details can be found @ https://www.cp2k.org/performance Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterCP2K Molecular Dynamics 2023.1Input: H2O-DFT-LSSMT OnSMT Off110022003300440055005012.264957.692143.512363.581. (F9X) gfortran options: -fopenmp -mtune=native -O3 -funroll-loops -fbacktrace -ffree-form -fimplicit-none -std=f2008 -lcp2kstart -lcp2kmc -lcp2kswarm -lcp2kmotion -lcp2kthermostat -lcp2kemd -lcp2ktmc -lcp2kmain -lcp2kdbt -lcp2ktas -lcp2kdbm -lcp2kgrid -lcp2kgridcpu -lcp2kgridref -lcp2kgridcommon -ldbcsrarnoldi -ldbcsrx -lcp2kshg_int -lcp2keri_mme -lcp2kminimax -lcp2khfxbase -lcp2ksubsys -lcp2kxc -lcp2kao -lcp2kpw_env -lcp2kinput -lcp2kpw -lcp2kgpu -lcp2kfft -lcp2kfpga -lcp2kfm -lcp2kcommon -lcp2koffload -lcp2kmpiwrap -lcp2kbase -ldbcsr -lsirius -lspla -lspfft -lsymspg -lvdwxc -lhdf5 -lhdf5_hl -lz -lgsl -lelpa_openmp -lcosma -lcosta -lscalapack -lxsmmf -lxsmm -ldl -lpthread -lxcf03 -lxc -lint2 -lfftw3_mpi -lfftw3 -lfftw3_omp -lmpi_cxx -lmpi -lopenblas -lvori -lstdc++ -lmpi_usempif08 -lmpi_mpifh -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterCP2K Molecular Dynamics 2023.1CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.44 / Avg: 291.52 / Max: 350.26Min: 20.12 / Avg: 290.26 / Max: 349.59Min: 41.56 / Avg: 614.44 / Max: 701.65Min: 40.71 / Avg: 623.68 / Max: 701.99

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 1 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path TracerSMT OnSMT Off2004006008001000SE +/- 0.88, N = 3SE +/- 1.20, N = 3SE +/- 2.85, N = 3SE +/- 0.58, N = 386111274826311. (CXX) g++ options: -O3 -lm -ldl
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOSPRay Studio 0.11CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.8 / Avg: 286.61 / Max: 335.12Min: 20.2 / Avg: 277.45 / Max: 323.23Min: 41.89 / Avg: 609 / Max: 686.18Min: 41.01 / Avg: 558.64 / Max: 651.98
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 1 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path TracerSMT OnSMT Off2004006008001000Min: 859 / Avg: 860.67 / Max: 862Min: 1125 / Avg: 1126.67 / Max: 1129Min: 479 / Avg: 482.33 / Max: 488Min: 630 / Avg: 631 / Max: 6321. (CXX) g++ options: -O3 -lm -ldl

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 2 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path TracerSMT OnSMT Off4K8K12K16K20KSE +/- 6.12, N = 3SE +/- 11.93, N = 3SE +/- 19.63, N = 3SE +/- 5.29, N = 313963182697817102321. (CXX) g++ options: -O3 -lm -ldl
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOSPRay Studio 0.11CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.65 / Avg: 305.15 / Max: 338.21Min: 10.61 / Avg: 295.55 / Max: 325.63Min: 43.88 / Avg: 615.69 / Max: 685.2Min: 41.78 / Avg: 589.05 / Max: 653.82
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 2 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path TracerSMT OnSMT Off3K6K9K12K15KMin: 13953 / Avg: 13962.67 / Max: 13974Min: 18247 / Avg: 18269 / Max: 18288Min: 7783 / Avg: 7816.67 / Max: 7851Min: 10224 / Avg: 10232 / Max: 102421. (CXX) g++ options: -O3 -lm -ldl

CloverLeaf

CloverLeaf is a Lagrangian-Eulerian hydrodynamics benchmark. This test profile currently makes use of CloverLeaf's OpenMP version and benchmarked with the clover_bm.in input file (Problem 5). Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterCloverLeafLagrangian-Eulerian HydrodynamicsSMT OnSMT Off510152025SE +/- 0.11, N = 4SE +/- 0.09, N = 5SE +/- 0.27, N = 4SE +/- 0.04, N = 412.009.2721.6515.151. (F9X) gfortran options: -O3 -march=native -funroll-loops -fopenmp
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterCloverLeafCPU Power Consumption MonitorSMT OnSMT Off70140210280350Min: 20.74 / Avg: 156.83 / Max: 210.59Min: 19.99 / Avg: 142.48 / Max: 202.51Min: 41.07 / Avg: 322.81 / Max: 382.87Min: 40.69 / Avg: 293.85 / Max: 373.73
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterCloverLeafLagrangian-Eulerian HydrodynamicsSMT OnSMT Off510152025Min: 11.82 / Avg: 12 / Max: 12.24Min: 9.1 / Avg: 9.27 / Max: 9.58Min: 21.1 / Avg: 21.65 / Max: 22.36Min: 15.04 / Avg: 15.15 / Max: 15.211. (F9X) gfortran options: -O3 -march=native -funroll-loops -fopenmp

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 3 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path TracerSMT OnSMT Off30060090012001500SE +/- 0.58, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.88, N = 3103213585827631. (CXX) g++ options: -O3 -lm -ldl
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOSPRay Studio 0.11CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.43 / Avg: 287.82 / Max: 335.59Min: 20.74 / Avg: 281.21 / Max: 323.91Min: 43.64 / Avg: 585.98 / Max: 682.76Min: 42.47 / Avg: 564.46 / Max: 651.55
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 3 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path TracerSMT OnSMT Off2004006008001000Min: 1031 / Avg: 1032 / Max: 1033Min: 1358 / Avg: 1358.33 / Max: 1359Min: 582 / Avg: 582.33 / Max: 583Min: 761 / Avg: 762.67 / Max: 7641. (CXX) g++ options: -O3 -lm -ldl

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 2 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path TracerSMT OnSMT Off8K16K24K32K40KSE +/- 21.53, N = 3SE +/- 53.62, N = 3SE +/- 41.40, N = 3SE +/- 96.56, N = 3278853653915731205451. (CXX) g++ options: -O3 -lm -ldl
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOSPRay Studio 0.11CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.44 / Avg: 307.82 / Max: 338.46Min: 20.77 / Avg: 284.22 / Max: 324.97Min: 44.88 / Avg: 575.16 / Max: 684.71Min: 42.15 / Avg: 573.14 / Max: 653.73
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 2 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path TracerSMT OnSMT Off6K12K18K24K30KMin: 27851 / Avg: 27885.33 / Max: 27925Min: 36451 / Avg: 36538.67 / Max: 36636Min: 15652 / Avg: 15731 / Max: 15792Min: 20359 / Avg: 20545 / Max: 206831. (CXX) g++ options: -O3 -lm -ldl

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: ChaCha20-Poly1305SMT OnSMT Off200000M400000M600000M800000M1000000MSE +/- 15599558.91, N = 3SE +/- 86138373.27, N = 3SE +/- 267574958.87, N = 3SE +/- 39882779.79, N = 34624158373203927826899879098173601107844895705231. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgbyte/s Per Watt, More Is BetterOpenSSL 3.1Algorithm: ChaCha20-Poly1305SMT OnSMT Off300M600M900M1200M1500M1284331827.021218755474.841256483723.091185929906.41
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenSSL 3.1CPU Power Consumption MonitorSMT OnSMT Off130260390520650Min: 21.26 / Avg: 360.04 / Max: 378.01Min: 20.77 / Avg: 322.28 / Max: 351.12Min: 42.98 / Avg: 724.1 / Max: 753.79Min: 42.18 / Avg: 661.5 / Max: 692.49
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: ChaCha20-Poly1305SMT OnSMT Off160000M320000M480000M640000M800000MMin: 462384997720 / Avg: 462415837320 / Max: 462435347110Min: 392631473770 / Avg: 392782689986.67 / Max: 392929780260Min: 909287160760 / Avg: 909817360110 / Max: 910145353050Min: 784424869180 / Avg: 784489570523.33 / Max: 7845623218201. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 2 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path TracerSMT OnSMT Off2004006008001000SE +/- 0.88, N = 3SE +/- 1.53, N = 3SE +/- 1.45, N = 3SE +/- 0.33, N = 387311464956451. (CXX) g++ options: -O3 -lm -ldl
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOSPRay Studio 0.11CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.25 / Avg: 287.03 / Max: 334.32Min: 20.75 / Avg: 278.21 / Max: 323.58Min: 44.18 / Avg: 586.29 / Max: 680.45Min: 42.03 / Avg: 559.72 / Max: 652.58
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 2 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path TracerSMT OnSMT Off2004006008001000Min: 872 / Avg: 873.33 / Max: 875Min: 1144 / Avg: 1146 / Max: 1149Min: 492 / Avg: 494.67 / Max: 497Min: 645 / Avg: 645.33 / Max: 6461. (CXX) g++ options: -O3 -lm -ldl

Graph500

This is a benchmark of the reference implementation of Graph500, an HPC benchmark focused on data intensive loads and commonly tested on supercomputers for complex data problems. Graph500 primarily stresses the communication subsystem of the hardware under test. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgsssp median_TEPS, More Is BetterGraph500 3.0Scale: 26SMT OnSMT Off160M320M480M640M800M3334450003637500006725410007701800001. (CC) gcc options: -fcommon -O3 -lpthread -lm -lmpi

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 1 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path TracerSMT OnSMT Off8K16K24K32K40KSE +/- 10.48, N = 3SE +/- 29.21, N = 3SE +/- 64.83, N = 3SE +/- 38.84, N = 3275383592615619201331. (CXX) g++ options: -O3 -lm -ldl
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOSPRay Studio 0.11CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.29 / Avg: 303.95 / Max: 338.98Min: 21.33 / Avg: 283.12 / Max: 324.52Min: 43.58 / Avg: 572.24 / Max: 686.68Min: 41.88 / Avg: 551.43 / Max: 653.67
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 1 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path TracerSMT OnSMT Off6K12K18K24K30KMin: 27525 / Avg: 27538.33 / Max: 27559Min: 35868 / Avg: 35926 / Max: 35961Min: 15520 / Avg: 15619 / Max: 15741Min: 20082 / Avg: 20132.67 / Max: 202091. (CXX) g++ options: -O3 -lm -ldl

Blender

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.6Blend File: Pabellon Barcelona - Compute: CPU-OnlySMT OnSMT Off1122334455SE +/- 0.08, N = 3SE +/- 0.14, N = 3SE +/- 0.08, N = 3SE +/- 0.09, N = 339.2750.0121.7527.47
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterBlender 3.6CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.55 / Avg: 303.74 / Max: 348.04Min: 21.11 / Avg: 297.19 / Max: 330.72Min: 43.07 / Avg: 552.47 / Max: 701.42Min: 194.9 / Avg: 571.12 / Max: 665.24
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.6Blend File: Pabellon Barcelona - Compute: CPU-OnlySMT OnSMT Off1020304050Min: 39.11 / Avg: 39.27 / Max: 39.36Min: 49.74 / Avg: 50.01 / Max: 50.2Min: 21.6 / Avg: 21.75 / Max: 21.84Min: 27.36 / Avg: 27.47 / Max: 27.64

John The Ripper

This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: WPA PSKSMT OnSMT Off300K600K900K1200K1500KSE +/- 505.98, N = 3SE +/- 6933.79, N = 3SE +/- 18163.51, N = 15SE +/- 15887.63, N = 4810375676218152453313015001. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgReal C/S Per Watt, More Is BetterJohn The Ripper 2023.03.14Test: WPA PSKSMT OnSMT Off50010001500200025002324.372240.312235.812047.36
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterJohn The Ripper 2023.03.14CPU Power Consumption MonitorSMT OnSMT Off140280420560700Min: 20.68 / Avg: 348.64 / Max: 397.25Min: 20.23 / Avg: 301.84 / Max: 349.66Min: 43.21 / Avg: 681.87 / Max: 792.38Min: 41.73 / Avg: 635.7 / Max: 734.09
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: WPA PSKSMT OnSMT Off300K600K900K1200K1500KMin: 809376 / Avg: 810375.33 / Max: 811013Min: 669013 / Avg: 676218 / Max: 690082Min: 1324000 / Avg: 1524533.33 / Max: 1572000Min: 1254000 / Avg: 1301500 / Max: 13210001. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt

Xmrig

Xmrig is an open-source cross-platform CPU/GPU miner for RandomX, KawPow, CryptoNight and AstroBWT. This test profile is setup to measure the Xmlrig CPU mining performance. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgH/s, More Is BetterXmrig 6.18.1Variant: Wownero - Hash Count: 1MSMT OnSMT Off30K60K90K120K150KSE +/- 513.11, N = 15SE +/- 13.77, N = 4SE +/- 677.82, N = 5SE +/- 741.13, N = 474803.663182.2142082.7100754.31. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgH/s Per Watt, More Is BetterXmrig 6.18.1Variant: Wownero - Hash Count: 1MSMT OnSMT Off70140210280350279.93235.83323.17225.36
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterXmrig 6.18.1CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.54 / Avg: 267.22 / Max: 350.81Min: 20.64 / Avg: 267.92 / Max: 342.62Min: 43.21 / Avg: 439.66 / Max: 708.61Min: 41.49 / Avg: 447.09 / Max: 663.47
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgH/s, More Is BetterXmrig 6.18.1Variant: Wownero - Hash Count: 1MSMT OnSMT Off20K40K60K80K100KMin: 73115.4 / Avg: 74803.56 / Max: 79846.7Min: 63159.2 / Avg: 63182.18 / Max: 63219.1Min: 139742.9 / Avg: 142082.68 / Max: 143760.8Min: 99364.1 / Avg: 100754.28 / Max: 102186.81. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.0Preset: ExhaustiveSMT OnSMT Off48121620SE +/- 0.0006, N = 5SE +/- 0.0028, N = 5SE +/- 0.0081, N = 6SE +/- 0.0257, N = 68.17437.314715.930514.40991. (CXX) g++ options: -O3 -flto -pthread
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgMT/s Per Watt, More Is BetterASTC Encoder 4.0Preset: ExhaustiveSMT OnSMT Off0.01260.02520.03780.05040.0630.0470.0420.0560.052
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterASTC Encoder 4.0CPU Power Consumption MonitorSMT OnSMT Off130260390520650Min: 19.98 / Avg: 172.1 / Max: 353.68Min: 19.61 / Avg: 174.24 / Max: 339.1Min: 41.13 / Avg: 283.59 / Max: 713.06Min: 40.28 / Avg: 278.8 / Max: 679.54
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.0Preset: ExhaustiveSMT OnSMT Off48121620Min: 8.17 / Avg: 8.17 / Max: 8.18Min: 7.31 / Avg: 7.31 / Max: 7.32Min: 15.91 / Avg: 15.93 / Max: 15.96Min: 14.34 / Avg: 14.41 / Max: 14.491. (CXX) g++ options: -O3 -flto -pthread

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Person Detection FP32 - Device: CPUSMT OnSMT Off5001000150020002500SE +/- 12.47, N = 15SE +/- 10.77, N = 5SE +/- 8.82, N = 3SE +/- 4.61, N = 32378.071102.821552.151579.141. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2022.3CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.83 / Avg: 304.46 / Max: 356.65Min: 20.58 / Avg: 307.41 / Max: 349.24Min: 44.3 / Avg: 527.63 / Max: 662.1Min: 195.17 / Avg: 529.98 / Max: 666.79
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Person Detection FP32 - Device: CPUSMT OnSMT Off400800120016002000Min: 2206.2 / Avg: 2378.07 / Max: 2404.93Min: 1060.16 / Avg: 1102.82 / Max: 1118.84Min: 1536.5 / Avg: 1552.15 / Max: 1567.04Min: 1573.46 / Avg: 1579.14 / Max: 1588.261. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Machine Translation EN To DE FP16 - Device: CPUSMT OnSMT Off30060090012001500SE +/- 6.51, N = 15SE +/- 4.46, N = 15SE +/- 7.44, N = 3SE +/- 4.51, N = 3582.08545.131159.991174.991. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Machine Translation EN To DE FP16 - Device: CPUSMT OnSMT Off2004006008001000Min: 564.62 / Avg: 582.08 / Max: 658Min: 531.41 / Avg: 545.13 / Max: 595.26Min: 1145.43 / Avg: 1159.99 / Max: 1169.89Min: 1167 / Avg: 1174.99 / Max: 1182.611. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Person Detection FP16 - Device: CPUSMT OnSMT Off5001000150020002500SE +/- 16.41, N = 12SE +/- 8.36, N = 9SE +/- 4.07, N = 3SE +/- 10.43, N = 32377.791105.961559.951545.231. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2022.3CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.09 / Avg: 304.04 / Max: 357.06Min: 20.69 / Avg: 306.74 / Max: 352.51Min: 43.22 / Avg: 524.02 / Max: 656.42Min: 199.54 / Avg: 537.68 / Max: 660.12
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Person Detection FP16 - Device: CPUSMT OnSMT Off400800120016002000Min: 2200.85 / Avg: 2377.79 / Max: 2413.38Min: 1044.6 / Avg: 1105.96 / Max: 1121.14Min: 1552.77 / Avg: 1559.95 / Max: 1566.85Min: 1525.11 / Avg: 1545.23 / Max: 1560.051. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

SPECFEM3D

simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.0Model: Layered HalfspaceSMT OnSMT Off48121620SE +/- 0.034780206, N = 3SE +/- 0.159428812, N = 3SE +/- 0.066481426, N = 15SE +/- 0.056354216, N = 415.96493168815.0281219299.9974475707.4707729301. (F9X) gfortran options: -O2 -fopenmp -std=f2003 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterSPECFEM3D 4.0CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.77 / Avg: 234.87 / Max: 334.16Min: 20.36 / Avg: 226.26 / Max: 322.57Min: 41.53 / Avg: 396.49 / Max: 677.63Min: 41.53 / Avg: 355.84 / Max: 655.14
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.0Model: Layered HalfspaceSMT OnSMT Off48121620Min: 15.9 / Avg: 15.96 / Max: 16.01Min: 14.71 / Avg: 15.03 / Max: 15.21Min: 9.58 / Avg: 10 / Max: 10.53Min: 7.4 / Avg: 7.47 / Max: 7.641. (F9X) gfortran options: -O2 -fopenmp -std=f2003 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

Blender

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.6Blend File: Barbershop - Compute: CPU-OnlySMT OnSMT Off306090120150SE +/- 0.11, N = 3SE +/- 0.17, N = 3SE +/- 0.19, N = 3SE +/- 0.08, N = 3116.54147.2269.0385.54
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterBlender 3.6CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.88 / Avg: 319.26 / Max: 346.14Min: 20.45 / Avg: 315.38 / Max: 335.91Min: 44.79 / Avg: 604.43 / Max: 701.89Min: 195.38 / Avg: 605.16 / Max: 674.35
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.6Blend File: Barbershop - Compute: CPU-OnlySMT OnSMT Off306090120150Min: 116.31 / Avg: 116.54 / Max: 116.67Min: 147.01 / Avg: 147.22 / Max: 147.56Min: 68.71 / Avg: 69.03 / Max: 69.38Min: 85.43 / Avg: 85.54 / Max: 85.69

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.6Blend File: BMW27 - Compute: CPU-OnlySMT OnSMT Off48121620SE +/- 0.02, N = 4SE +/- 0.05, N = 4SE +/- 0.02, N = 6SE +/- 0.05, N = 512.7715.157.128.45
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterBlender 3.6CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.06 / Avg: 242.87 / Max: 337.46Min: 20.52 / Avg: 245.21 / Max: 323.25Min: 42.04 / Avg: 405.53 / Max: 681.36Min: 196.82 / Avg: 459.78 / Max: 647.78
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.6Blend File: BMW27 - Compute: CPU-OnlySMT OnSMT Off48121620Min: 12.72 / Avg: 12.77 / Max: 12.82Min: 15.05 / Avg: 15.15 / Max: 15.27Min: 7.06 / Avg: 7.12 / Max: 7.17Min: 8.36 / Avg: 8.45 / Max: 8.66

Helsing

Helsing is an open-source POSIX vampire number generator. This test profile measures the time it takes to generate vampire numbers between varying numbers of digits. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterHelsing 1.0-betaDigit Range: 14 digitSMT OnSMT Off1326395265SE +/- 0.09, N = 3SE +/- 0.03, N = 3SE +/- 0.37, N = 3SE +/- 0.31, N = 350.4757.9627.2850.051. (CC) gcc options: -O2 -pthread
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterHelsing 1.0-betaCPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.03 / Avg: 316.26 / Max: 355.5Min: 20.68 / Avg: 310.21 / Max: 356.64Min: 42.81 / Avg: 589.87 / Max: 707.99Min: 41.78 / Avg: 512.26 / Max: 690.1
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterHelsing 1.0-betaDigit Range: 14 digitSMT OnSMT Off1122334455Min: 50.33 / Avg: 50.47 / Max: 50.63Min: 57.91 / Avg: 57.96 / Max: 58.02Min: 26.82 / Avg: 27.28 / Max: 28Min: 49.46 / Avg: 50.05 / Max: 50.51. (CC) gcc options: -O2 -pthread

Graph500

This is a benchmark of the reference implementation of Graph500, an HPC benchmark focused on data intensive loads and commonly tested on supercomputers for complex data problems. Graph500 primarily stresses the communication subsystem of the hardware under test. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgbfs median_TEPS, More Is BetterGraph500 3.0Scale: 26SMT OnSMT Off400M800M1200M1600M2000M857890000893624000157110000018151600001. (CC) gcc options: -fcommon -O3 -lpthread -lm -lmpi

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgverify/s, More Is BetterOpenSSL 3.1Algorithm: RSA4096SMT OnSMT Off800K1600K2400K3200K4000KSE +/- 405.88, N = 3SE +/- 1031.37, N = 3SE +/- 163.02, N = 3SE +/- 1067.40, N = 31890935.31799293.03782091.83598946.31. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgverify/s Per Watt, More Is BetterOpenSSL 3.1Algorithm: RSA4096SMT OnSMT Off140028004200560070006020.726318.786122.576002.69
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenSSL 3.1CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.6 / Avg: 314.07 / Max: 354.72Min: 21.32 / Avg: 284.75 / Max: 316.61Min: 44.51 / Avg: 617.73 / Max: 697.4Min: 42.69 / Avg: 599.56 / Max: 681.84
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgverify/s, More Is BetterOpenSSL 3.1Algorithm: RSA4096SMT OnSMT Off700K1400K2100K2800K3500KMin: 1890303.7 / Avg: 1890935.27 / Max: 1891692.7Min: 1797448.2 / Avg: 1799293.03 / Max: 1801014.6Min: 3781915.6 / Avg: 3782091.83 / Max: 3782417.5Min: 3597468.2 / Avg: 3598946.3 / Max: 3601019.31. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: MG.CSMT OnSMT Off60K120K180K240K300KSE +/- 296.36, N = 10SE +/- 104.98, N = 10SE +/- 1284.96, N = 11SE +/- 599.41, N = 10128129.56136942.13249109.09268721.051. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s Per Watt, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: MG.CSMT OnSMT Off300600900120015001449.061502.141433.881554.08
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNAS Parallel Benchmarks 3.4CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 20.16 / Avg: 88.42 / Max: 270.94Min: 19.66 / Avg: 91.16 / Max: 281.3Min: 41.54 / Avg: 173.73 / Max: 526.19Min: 40.59 / Avg: 172.91 / Max: 610.41
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: MG.CSMT OnSMT Off50K100K150K200K250KMin: 126992.55 / Avg: 128129.56 / Max: 130078.64Min: 136114.6 / Avg: 136942.13 / Max: 137303.36Min: 243709.43 / Avg: 249109.09 / Max: 255904.32Min: 264873.03 / Avg: 268721.05 / Max: 270846.391. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.0Preset: FastSMT OnSMT Off30060090012001500SE +/- 1.79, N = 6SE +/- 1.20, N = 7SE +/- 1.29, N = 5SE +/- 1.25, N = 51190.751278.65610.11693.401. (CXX) g++ options: -O3 -flto -pthread
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgMT/s Per Watt, More Is BetterASTC Encoder 4.0Preset: FastSMT OnSMT Off36912159.67110.8982.4702.954
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterASTC Encoder 4.0CPU Power Consumption MonitorSMT OnSMT Off80160240320400Min: 20.41 / Avg: 123.13 / Max: 316.82Min: 20.06 / Avg: 117.33 / Max: 315.16Min: 41.99 / Avg: 247.05 / Max: 432.29Min: 41.7 / Avg: 234.69 / Max: 432.58
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.0Preset: FastSMT OnSMT Off2004006008001000Min: 1187.12 / Avg: 1190.75 / Max: 1198.62Min: 1274.28 / Avg: 1278.65 / Max: 1282.12Min: 605.83 / Avg: 610.11 / Max: 613.09Min: 689.68 / Avg: 693.4 / Max: 697.111. (CXX) g++ options: -O3 -flto -pthread

OSPRay

Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 2.12Benchmark: gravity_spheres_volume/dim_512/ao/real_timeSMT OnSMT Off1224364860SE +/- 0.07, N = 3SE +/- 0.01, N = 3SE +/- 0.09, N = 3SE +/- 0.02, N = 332.6825.7253.8344.71
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgItems Per Second Per Watt, More Is BetterOSPRay 2.12Benchmark: gravity_spheres_volume/dim_512/ao/real_timeSMT OnSMT Off0.02660.05320.07980.10640.1330.1180.1010.0950.082
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOSPRay 2.12CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.45 / Avg: 276.76 / Max: 329.91Min: 20.16 / Avg: 254.24 / Max: 300.93Min: 44.15 / Avg: 566.91 / Max: 661.42Min: 40.66 / Avg: 542.85 / Max: 624.55
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 2.12Benchmark: gravity_spheres_volume/dim_512/ao/real_timeSMT OnSMT Off1122334455Min: 32.55 / Avg: 32.68 / Max: 32.78Min: 25.7 / Avg: 25.72 / Max: 25.74Min: 53.65 / Avg: 53.83 / Max: 53.93Min: 44.68 / Avg: 44.71 / Max: 44.76

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgsign/s, More Is BetterOpenSSL 3.1Algorithm: RSA4096SMT OnSMT Off20K40K60K80K100KSE +/- 16.66, N = 3SE +/- 1.80, N = 3SE +/- 3.93, N = 3SE +/- 3.85, N = 354195.156647.3108490.5113251.41. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgsign/s, More Is BetterOpenSSL 3.1Algorithm: RSA4096SMT OnSMT Off20K40K60K80K100KMin: 54175.2 / Avg: 54195.1 / Max: 54228.2Min: 56643.7 / Avg: 56647.3 / Max: 56649.3Min: 108486 / Avg: 108490.47 / Max: 108498.3Min: 113243.8 / Avg: 113251.43 / Max: 113256.11. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

John The Ripper

This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: MD5SMT OnSMT Off7M14M21M28M35MSE +/- 52818.35, N = 3SE +/- 35950.58, N = 3SE +/- 83819.91, N = 3SE +/- 140717.61, N = 3203126671675166734879333302216671. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgReal C/S Per Watt, More Is BetterJohn The Ripper 2023.03.14Test: MD5SMT OnSMT Off15K30K45K60K75K69625.9056420.6367440.8057347.31
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterJohn The Ripper 2023.03.14CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.42 / Avg: 291.74 / Max: 363.73Min: 20.4 / Avg: 296.91 / Max: 362.1Min: 41.25 / Avg: 517.18 / Max: 684.78Min: 40.66 / Avg: 526.99 / Max: 698.52
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: MD5SMT OnSMT Off6M12M18M24M30MMin: 20242000 / Avg: 20312666.67 / Max: 20416000Min: 16693000 / Avg: 16751666.67 / Max: 16817000Min: 34714000 / Avg: 34879333.33 / Max: 34986000Min: 29941000 / Avg: 30221666.67 / Max: 303800001. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt

OSPRay

Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 2.12Benchmark: particle_volume/scivis/real_timeSMT OnSMT Off1122334455SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 330.8223.7249.2341.55
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgItems Per Second Per Watt, More Is BetterOSPRay 2.12Benchmark: particle_volume/scivis/real_timeSMT OnSMT Off0.02860.05720.08580.11440.1430.1270.1120.1020.090
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOSPRay 2.12CPU Power Consumption MonitorSMT OnSMT Off100200300400500Min: 20.96 / Avg: 241.99 / Max: 304.45Min: 20.16 / Avg: 211.09 / Max: 275.73Min: 44 / Avg: 483.3 / Max: 582.81Min: 42.19 / Avg: 462.03 / Max: 561.22
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 2.12Benchmark: particle_volume/scivis/real_timeSMT OnSMT Off1020304050Min: 30.79 / Avg: 30.82 / Max: 30.84Min: 23.69 / Avg: 23.72 / Max: 23.74Min: 49.19 / Avg: 49.23 / Max: 49.3Min: 41.5 / Avg: 41.55 / Max: 41.59

Blender

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.6Blend File: Fishy Cat - Compute: CPU-OnlySMT OnSMT Off510152025SE +/- 0.06, N = 3SE +/- 0.06, N = 3SE +/- 0.02, N = 5SE +/- 0.04, N = 416.4920.479.8711.72
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterBlender 3.6CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.92 / Avg: 259.04 / Max: 346.77Min: 20.68 / Avg: 257.96 / Max: 331.22Min: 42.57 / Avg: 439.22 / Max: 700.81Min: 194.02 / Avg: 484.54 / Max: 664
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.6Blend File: Fishy Cat - Compute: CPU-OnlySMT OnSMT Off510152025Min: 16.38 / Avg: 16.49 / Max: 16.6Min: 20.37 / Avg: 20.47 / Max: 20.56Min: 9.81 / Avg: 9.87 / Max: 9.95Min: 11.63 / Avg: 11.72 / Max: 11.84

OSPRay

Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 2.12Benchmark: particle_volume/ao/real_timeSMT OnSMT Off1122334455SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 330.8523.7349.1341.66
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgItems Per Second Per Watt, More Is BetterOSPRay 2.12Benchmark: particle_volume/ao/real_timeSMT OnSMT Off0.02390.04780.07170.09560.11950.1060.0910.0890.078
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOSPRay 2.12CPU Power Consumption MonitorSMT OnSMT Off100200300400500Min: 21.03 / Avg: 291.76 / Max: 307.53Min: 20.26 / Avg: 261.53 / Max: 278.17Min: 41.99 / Avg: 553.35 / Max: 585.29Min: 42.16 / Avg: 534.35 / Max: 564.38
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 2.12Benchmark: particle_volume/ao/real_timeSMT OnSMT Off1020304050Min: 30.8 / Avg: 30.85 / Max: 30.9Min: 23.7 / Avg: 23.73 / Max: 23.76Min: 49.06 / Avg: 49.13 / Max: 49.23Min: 41.62 / Avg: 41.66 / Max: 41.69

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 2.12Benchmark: gravity_spheres_volume/dim_512/scivis/real_timeSMT OnSMT Off1224364860SE +/- 0.09, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.07, N = 331.9225.5152.7344.29
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgItems Per Second Per Watt, More Is BetterOSPRay 2.12Benchmark: gravity_spheres_volume/dim_512/scivis/real_timeSMT OnSMT Off0.02590.05180.07770.10360.12950.1150.1000.0920.081
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOSPRay 2.12CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.29 / Avg: 278.61 / Max: 330.56Min: 20.36 / Avg: 255.54 / Max: 301.79Min: 42.48 / Avg: 572.66 / Max: 664.61Min: 41.49 / Avg: 544.27 / Max: 626.51
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 2.12Benchmark: gravity_spheres_volume/dim_512/scivis/real_timeSMT OnSMT Off1122334455Min: 31.75 / Avg: 31.92 / Max: 32.02Min: 25.46 / Avg: 25.51 / Max: 25.54Min: 52.68 / Avg: 52.73 / Max: 52.77Min: 44.18 / Avg: 44.29 / Max: 44.42

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: SHA512SMT OnSMT Off20000M40000M60000M80000M100000MSE +/- 4276543.39, N = 3SE +/- 19506083.54, N = 3SE +/- 22577791.79, N = 3SE +/- 660141733.06, N = 353005879330518043336371060496552031022322335901. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgbyte/s Per Watt, More Is BetterOpenSSL 3.1Algorithm: SHA512SMT OnSMT Off30M60M90M120M150M157871758.84155713911.66156820493.54150591857.13
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenSSL 3.1CPU Power Consumption MonitorSMT OnSMT Off130260390520650Min: 21.29 / Avg: 335.75 / Max: 346.67Min: 20.38 / Avg: 332.69 / Max: 354.84Min: 44.01 / Avg: 676.25 / Max: 700.91Min: 41.9 / Avg: 678.87 / Max: 711.36
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: SHA512SMT OnSMT Off20000M40000M60000M80000M100000MMin: 53000303950 / Avg: 53005879330 / Max: 53014284210Min: 51772795230 / Avg: 51804333636.67 / Max: 51839988980Min: 106004564450 / Avg: 106049655203.33 / Max: 106074295340Min: 101383737990 / Avg: 102232233590 / Max: 1035324966501. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Face Detection FP16 - Device: CPUSMT OnSMT Off2004006008001000SE +/- 0.05, N = 3SE +/- 0.07, N = 3SE +/- 0.18, N = 3SE +/- 0.28, N = 31049.31513.23526.98526.461. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2022.3CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.39 / Avg: 315.89 / Max: 357.76Min: 20.94 / Avg: 316.32 / Max: 349.13Min: 45.01 / Avg: 611.1 / Max: 703.01Min: 193.21 / Avg: 621.51 / Max: 702.85
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Face Detection FP16 - Device: CPUSMT OnSMT Off2004006008001000Min: 1049.24 / Avg: 1049.31 / Max: 1049.41Min: 513.09 / Avg: 513.23 / Max: 513.33Min: 526.63 / Avg: 526.98 / Max: 527.24Min: 525.91 / Avg: 526.46 / Max: 526.811. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: AES-128-GCMSMT OnSMT Off500000M1000000M1500000M2000000M2500000MSE +/- 404585301.69, N = 3SE +/- 772883363.35, N = 3SE +/- 4494053056.14, N = 3SE +/- 4718005576.14, N = 311696737355571152838928063233989070010723075245354571. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgbyte/s Per Watt, More Is BetterOpenSSL 3.1Algorithm: AES-128-GCMSMT OnSMT Off700M1400M2100M2800M3500M3408107334.733373392798.473375379415.533357766493.21
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenSSL 3.1CPU Power Consumption MonitorSMT OnSMT Off130260390520650Min: 20.93 / Avg: 343.2 / Max: 354.82Min: 20.8 / Avg: 341.74 / Max: 355.14Min: 42.1 / Avg: 693.22 / Max: 715.73Min: 42.47 / Avg: 687.22 / Max: 716.34
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: AES-128-GCMSMT OnSMT Off400000M800000M1200000M1600000M2000000MMin: 1168870904080 / Avg: 1169673735556.7 / Max: 1170162695650Min: 1151848296170 / Avg: 1152838928063.3 / Max: 1154361878390Min: 2331034879590 / Avg: 2339890700106.7 / Max: 2345649163200Min: 2298662876500 / Avg: 2307524535456.7 / Max: 23147628514001. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: AES-256-GCMSMT OnSMT Off400000M800000M1200000M1600000M2000000MSE +/- 2018584981.53, N = 3SE +/- 917614258.55, N = 3SE +/- 676471418.80, N = 3SE +/- 2904032106.06, N = 31012537766210997034621883201931707032719982341488631. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgbyte/s Per Watt, More Is BetterOpenSSL 3.1Algorithm: AES-256-GCMSMT OnSMT Off600M1200M1800M2400M3000M2960716545.712959773610.692919004979.802960156985.81
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenSSL 3.1CPU Power Consumption MonitorSMT OnSMT Off130260390520650Min: 21.31 / Avg: 341.99 / Max: 354.9Min: 20.86 / Avg: 336.86 / Max: 352.15Min: 45.1 / Avg: 691.78 / Max: 716.83Min: 42.53 / Avg: 675.04 / Max: 707.9
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: AES-256-GCMSMT OnSMT Off400000M800000M1200000M1600000M2000000MMin: 1008505646280 / Avg: 1012537766210 / Max: 1014728647900Min: 995569430250 / Avg: 997034621883.33 / Max: 998724274590Min: 2018497738890 / Avg: 2019317070326.7 / Max: 2020659132280Min: 1993569074650 / Avg: 1998234148863.3 / Max: 20035630202901. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Machine Translation EN To DE FP16 - Device: CPUSMT OnSMT Off20406080100SE +/- 1.12, N = 15SE +/- 0.45, N = 15SE +/- 0.35, N = 3SE +/- 0.21, N = 3110.0058.7155.1254.421. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2022.3CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.17 / Avg: 293.93 / Max: 343.74Min: 20.68 / Avg: 303.27 / Max: 334.77Min: 44.56 / Avg: 611.86 / Max: 676.4Min: 197.64 / Avg: 618.57 / Max: 675.82
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Machine Translation EN To DE FP16 - Device: CPUSMT OnSMT Off20406080100Min: 97.18 / Avg: 110 / Max: 113.23Min: 53.72 / Avg: 58.71 / Max: 60.17Min: 54.65 / Avg: 55.12 / Max: 55.81Min: 54.07 / Avg: 54.42 / Max: 54.791. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Face Detection FP16-INT8 - Device: CPUSMT OnSMT Off120240360480600SE +/- 0.09, N = 3SE +/- 0.06, N = 3SE +/- 0.07, N = 3SE +/- 0.06, N = 3541.29268.82270.93271.241. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2022.3CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.62 / Avg: 295.03 / Max: 335.3Min: 20.4 / Avg: 294.88 / Max: 329.14Min: 45.43 / Avg: 578.64 / Max: 665.53Min: 195.77 / Avg: 585.67 / Max: 664.94
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Face Detection FP16-INT8 - Device: CPUSMT OnSMT Off100200300400500Min: 541.12 / Avg: 541.29 / Max: 541.41Min: 268.71 / Avg: 268.82 / Max: 268.89Min: 270.8 / Avg: 270.93 / Max: 271.01Min: 271.14 / Avg: 271.24 / Max: 271.351. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Face Detection FP16-INT8 - Device: CPUSMT OnSMT Off50100150200250SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.09, N = 3SE +/- 0.08, N = 3117.88119.01235.65235.371. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Face Detection FP16-INT8 - Device: CPUSMT OnSMT Off4080120160200Min: 117.82 / Avg: 117.88 / Max: 117.95Min: 118.98 / Avg: 119.01 / Max: 119.05Min: 235.49 / Avg: 235.65 / Max: 235.81Min: 235.27 / Avg: 235.37 / Max: 235.541. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Face Detection FP16 - Device: CPUSMT OnSMT Off306090120150SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 360.7762.18121.03121.081. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Face Detection FP16 - Device: CPUSMT OnSMT Off20406080100Min: 60.74 / Avg: 60.77 / Max: 60.81Min: 62.17 / Avg: 62.18 / Max: 62.19Min: 120.96 / Avg: 121.03 / Max: 121.12Min: 121.06 / Avg: 121.08 / Max: 121.121. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-StreamSMT OnSMT Off6001200180024003000SE +/- 1.21, N = 3SE +/- 3.84, N = 3SE +/- 6.95, N = 3SE +/- 33.47, N = 151376.211780.442627.312730.50
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-StreamSMT OnSMT Off5001000150020002500Min: 1374.35 / Avg: 1376.21 / Max: 1378.5Min: 1774.35 / Avg: 1780.44 / Max: 1787.54Min: 2616.31 / Avg: 2627.31 / Max: 2640.16Min: 2497.03 / Avg: 2730.5 / Max: 2961.2

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.0Preset: ThoroughSMT OnSMT Off306090120150SE +/- 0.02, N = 6SE +/- 0.01, N = 6SE +/- 0.22, N = 6SE +/- 0.05, N = 675.0868.31134.89127.211. (CXX) g++ options: -O3 -flto -pthread
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgMT/s Per Watt, More Is BetterASTC Encoder 4.0Preset: ThoroughSMT OnSMT Off0.13320.26640.39960.53280.6660.5610.5070.5920.577
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterASTC Encoder 4.0CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 14.08 / Avg: 133.8 / Max: 351.54Min: 19.58 / Avg: 134.61 / Max: 338.88Min: 41.01 / Avg: 227.98 / Max: 671.57Min: 40.08 / Avg: 220.5 / Max: 656.14
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.0Preset: ThoroughSMT OnSMT Off306090120150Min: 75.01 / Avg: 75.08 / Max: 75.18Min: 68.27 / Avg: 68.31 / Max: 68.35Min: 134.29 / Avg: 134.89 / Max: 135.73Min: 127.04 / Avg: 127.21 / Max: 127.351. (CXX) g++ options: -O3 -flto -pthread

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Weld Porosity Detection FP16-INT8 - Device: CPUSMT OnSMT Off5K10K15K20K25KSE +/- 2.13, N = 3SE +/- 13.11, N = 3SE +/- 16.03, N = 3SE +/- 15.58, N = 311794.3211710.5922954.4422955.641. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Weld Porosity Detection FP16-INT8 - Device: CPUSMT OnSMT Off4K8K12K16K20KMin: 11791.82 / Avg: 11794.32 / Max: 11798.56Min: 11697.14 / Avg: 11710.59 / Max: 11736.81Min: 22922.71 / Avg: 22954.44 / Max: 22974.28Min: 22924.48 / Avg: 22955.64 / Max: 22971.731. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Weld Porosity Detection FP16 - Device: CPUSMT OnSMT Off2K4K6K8K10KSE +/- 0.58, N = 3SE +/- 11.25, N = 3SE +/- 9.34, N = 3SE +/- 2.58, N = 36067.785837.6211299.3211373.951. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Weld Porosity Detection FP16 - Device: CPUSMT OnSMT Off2K4K6K8K10KMin: 6066.62 / Avg: 6067.78 / Max: 6068.49Min: 5818.78 / Avg: 5837.62 / Max: 5857.7Min: 11281.81 / Avg: 11299.32 / Max: 11313.73Min: 11370.76 / Avg: 11373.95 / Max: 11379.051. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

NAMD

NAMD is a parallel molecular dynamics code designed for high-performance simulation of large biomolecular systems. NAMD was developed by the Theoretical and Computational Biophysics Group in the Beckman Institute for Advanced Science and Technology at the University of Illinois at Urbana-Champaign. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgdays/ns, Fewer Is BetterNAMD 2.14ATPase Simulation - 327,506 AtomsSMT OnSMT Off0.04660.09320.13980.18640.233SE +/- 0.00095, N = 4SE +/- 0.00018, N = 4SE +/- 0.00040, N = 3SE +/- 0.00135, N = 50.207020.205950.106460.13969
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNAMD 2.14CPU Power Consumption MonitorSMT OnSMT Off130260390520650Min: 21.34 / Avg: 258.84 / Max: 353.57Min: 20.91 / Avg: 242.43 / Max: 333.11Min: 43.91 / Avg: 386.17 / Max: 712.38Min: 42.59 / Avg: 421.91 / Max: 639.89
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgdays/ns, Fewer Is BetterNAMD 2.14ATPase Simulation - 327,506 AtomsSMT OnSMT Off12345Min: 0.21 / Avg: 0.21 / Max: 0.21Min: 0.21 / Avg: 0.21 / Max: 0.21Min: 0.11 / Avg: 0.11 / Max: 0.11Min: 0.14 / Avg: 0.14 / Max: 0.14

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 256 - Buffer Length: 256 - Filter Length: 512SMT OnSMT Off500M1000M1500M2000M2500MSE +/- 470224.53, N = 3SE +/- 592546.29, N = 3SE +/- 1877054.43, N = 3SE +/- 1068228.02, N = 316967666671313933333254440000025426333331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgsamples/s Per Watt, More Is BetterLiquid-DSP 1.6Threads: 256 - Buffer Length: 256 - Filter Length: 512SMT OnSMT Off1.3M2.6M3.9M5.2M6.5M6258723.145048094.724829201.804850824.91
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterLiquid-DSP 1.6CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 20.84 / Avg: 271.1 / Max: 312.91Min: 20.66 / Avg: 260.28 / Max: 300.73Min: 42.35 / Avg: 526.88 / Max: 607.73Min: 41.3 / Avg: 524.17 / Max: 605.41
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 256 - Buffer Length: 256 - Filter Length: 512SMT OnSMT Off400M800M1200M1600M2000MMin: 1696200000 / Avg: 1696766666.67 / Max: 1697700000Min: 1312800000 / Avg: 1313933333.33 / Max: 1314800000Min: 2541100000 / Avg: 2544400000 / Max: 2547600000Min: 2540500000 / Avg: 2542633333.33 / Max: 25438000001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Person Vehicle Bike Detection FP16 - Device: CPUSMT OnSMT Off2K4K6K8K10KSE +/- 48.86, N = 15SE +/- 5.00, N = 3SE +/- 11.31, N = 3SE +/- 10.15, N = 36148.895118.299889.619878.881. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Person Vehicle Bike Detection FP16 - Device: CPUSMT OnSMT Off2K4K6K8K10KMin: 5963.92 / Avg: 6148.89 / Max: 6579.24Min: 5109.54 / Avg: 5118.29 / Max: 5126.86Min: 9867.93 / Avg: 9889.61 / Max: 9906.05Min: 9858.65 / Avg: 9878.88 / Max: 9890.461. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

Timed Linux Kernel Compilation

This test times how long it takes to build the Linux kernel in a default configuration (defconfig) for the architecture being tested or alternatively an allmodconfig for building all possible kernel modules for the build. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.1Build: allmodconfigSMT OnSMT Off50100150200250SE +/- 1.35, N = 3SE +/- 0.48, N = 3SE +/- 0.51, N = 3SE +/- 0.55, N = 3227.52177.76145.93118.10
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterTimed Linux Kernel Compilation 6.1CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.05 / Avg: 255.95 / Max: 334.41Min: 20.38 / Avg: 265.69 / Max: 337.05Min: 41.29 / Avg: 468.11 / Max: 673.75Min: 40.59 / Avg: 462.79 / Max: 676.08
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.1Build: allmodconfigSMT OnSMT Off4080120160200Min: 225.91 / Avg: 227.52 / Max: 230.2Min: 177.26 / Avg: 177.76 / Max: 178.73Min: 145.14 / Avg: 145.93 / Max: 146.89Min: 117.44 / Avg: 118.1 / Max: 119.18

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Weld Porosity Detection FP16 - Device: CPUSMT OnSMT Off3691215SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 310.535.475.645.611. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2022.3CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.7 / Avg: 327.23 / Max: 352.86Min: 20.96 / Avg: 318.56 / Max: 343.94Min: 44.97 / Avg: 643.58 / Max: 697.31Min: 197.92 / Avg: 655.04 / Max: 695.42
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Weld Porosity Detection FP16 - Device: CPUSMT OnSMT Off3691215Min: 10.53 / Avg: 10.53 / Max: 10.53Min: 5.45 / Avg: 5.47 / Max: 5.49Min: 5.64 / Avg: 5.64 / Max: 5.65Min: 5.61 / Avg: 5.61 / Max: 5.611. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

Primesieve

Primesieve generates prime numbers using a highly optimized sieve of Eratosthenes implementation. Primesieve primarily benchmarks the CPU's L1/L2 cache performance. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterPrimesieve 8.0Length: 1e13SMT OnSMT Off510152025SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 5SE +/- 0.03, N = 521.2921.1311.1211.151. (CXX) g++ options: -O3
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterPrimesieve 8.0CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.09 / Avg: 268.37 / Max: 324.15Min: 19.88 / Avg: 261.85 / Max: 316.4Min: 41.84 / Avg: 485.63 / Max: 670.4Min: 41.23 / Avg: 464.71 / Max: 636.03
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterPrimesieve 8.0Length: 1e13SMT OnSMT Off510152025Min: 21.24 / Avg: 21.29 / Max: 21.31Min: 21.1 / Avg: 21.13 / Max: 21.17Min: 11.07 / Avg: 11.12 / Max: 11.15Min: 11.07 / Avg: 11.15 / Max: 11.261. (CXX) g++ options: -O3

libxsmm

Libxsmm is an open-source library for specialized dense and sparse matrix operations and deep learning primitives. Libxsmm supports making use of Intel AMX, AVX-512, and other modern CPU instruction set capabilities. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOPS/s, More Is Betterlibxsmm 2-1.17-3645M N K: 256SMT OnSMT Off14002800420056007000SE +/- 2.32, N = 3SE +/- 16.75, N = 3SE +/- 1.43, N = 3SE +/- 66.66, N = 93331.73813.46112.66373.01. (CXX) g++ options: -dynamic -Bstatic -static-libgcc -lgomp -lm -lrt -ldl -lquadmath -lstdc++ -pthread -fPIC -std=c++14 -O2 -fopenmp-simd -funroll-loops -ftree-vectorize -fdata-sections -ffunction-sections -fvisibility=hidden -msse4.2
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOPS/s Per Watt, More Is Betterlibxsmm 2-1.17-3645M N K: 256SMT OnSMT Off51015202515.7818.1219.0419.92
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is Betterlibxsmm 2-1.17-3645CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.42 / Avg: 211.17 / Max: 339.24Min: 20.05 / Avg: 210.44 / Max: 332.07Min: 41.94 / Avg: 321.06 / Max: 680.36Min: 40.86 / Avg: 319.88 / Max: 674.35
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOPS/s, More Is Betterlibxsmm 2-1.17-3645M N K: 256SMT OnSMT Off11002200330044005500Min: 3328.9 / Avg: 3331.7 / Max: 3336.3Min: 3795.3 / Avg: 3813.43 / Max: 3846.9Min: 6111.2 / Avg: 6112.63 / Max: 6115.5Min: 6057.6 / Avg: 6372.96 / Max: 6613.51. (CXX) g++ options: -dynamic -Bstatic -static-libgcc -lgomp -lm -lrt -ldl -lquadmath -lstdc++ -pthread -fPIC -std=c++14 -O2 -fopenmp-simd -funroll-loops -ftree-vectorize -fdata-sections -ffunction-sections -fvisibility=hidden -msse4.2

toyBrot Fractal Generator

ToyBrot is a Mandelbrot fractal generator supporting C++ threads/tasks, OpenMP, Intel Threaded Building Blocks (TBB), and other targets. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BettertoyBrot Fractal Generator 2020-11-18Implementation: OpenMPSMT OnSMT Off13002600390052006500SE +/- 17.08, N = 8SE +/- 0.20, N = 7SE +/- 23.29, N = 12SE +/- 53.03, N = 1540816242332136711. (CXX) g++ options: -O3 -lpthread -lm -lgcc -lgcc_s -lc
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BettertoyBrot Fractal Generator 2020-11-18CPU Power Consumption MonitorSMT OnSMT Off100200300400500Min: 20.45 / Avg: 148.84 / Max: 302.49Min: 19.84 / Avg: 154.08 / Max: 274.28Min: 40.87 / Avg: 242.46 / Max: 590.56Min: 40.79 / Avg: 251.07 / Max: 574.98
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BettertoyBrot Fractal Generator 2020-11-18Implementation: OpenMPSMT OnSMT Off11002200330044005500Min: 4010 / Avg: 4081.13 / Max: 4149Min: 6241 / Avg: 6241.57 / Max: 6242Min: 3174 / Avg: 3320.58 / Max: 3479Min: 3486 / Avg: 3671.33 / Max: 41451. (CXX) g++ options: -O3 -lpthread -lm -lgcc -lgcc_s -lc

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPUSMT OnSMT Off30K60K90K120K150KSE +/- 192.97, N = 3SE +/- 126.57, N = 3SE +/- 602.14, N = 3SE +/- 420.06, N = 385400.8871673.85133931.53118225.551. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPUSMT OnSMT Off20K40K60K80K100KMin: 85165.48 / Avg: 85400.88 / Max: 85783.43Min: 71441.15 / Avg: 71673.85 / Max: 71876.48Min: 132727.25 / Avg: 133931.53 / Max: 134536.03Min: 117492.79 / Avg: 118225.55 / Max: 118947.821. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

miniBUDE

MiniBUDE is a mini application for the the core computation of the Bristol University Docking Engine (BUDE). This test profile currently makes use of the OpenMP implementation of miniBUDE for CPU benchmarking. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFInst/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM2SMT OnSMT Off2K4K6K8K10KSE +/- 0.21, N = 3SE +/- 6.13, N = 3SE +/- 10.75, N = 4SE +/- 150.60, N = 125972.645903.217888.0910989.941. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFInst/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM2SMT OnSMT Off2K4K6K8K10KMin: 5972.24 / Avg: 5972.64 / Max: 5972.96Min: 5891.56 / Avg: 5903.21 / Max: 5912.33Min: 7863.07 / Avg: 7888.08 / Max: 7911.53Min: 10135.99 / Avg: 10989.94 / Max: 11597.731. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgBillion Interactions/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM2SMT OnSMT Off100200300400500SE +/- 0.01, N = 3SE +/- 0.25, N = 3SE +/- 0.43, N = 4SE +/- 6.02, N = 12238.91236.13315.52439.601. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgBillion Interactions/s Per Watt, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM2SMT OnSMT Off0.21350.4270.64050.8541.06750.8830.8810.8190.949
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterminiBUDE 20210901CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 20.25 / Avg: 270.52 / Max: 323.64Min: 19.67 / Avg: 268.1 / Max: 319.7Min: 40.87 / Avg: 385.07 / Max: 627.6Min: 39.97 / Avg: 463.17 / Max: 639.3
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgBillion Interactions/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM2SMT OnSMT Off80160240320400Min: 238.89 / Avg: 238.91 / Max: 238.92Min: 235.66 / Avg: 236.13 / Max: 236.49Min: 314.52 / Avg: 315.52 / Max: 316.46Min: 405.44 / Avg: 439.6 / Max: 463.911. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: IS.DSMT OnSMT Off2K4K6K8K10KSE +/- 27.01, N = 6SE +/- 29.88, N = 6SE +/- 104.60, N = 15SE +/- 105.10, N = 155300.295315.159849.018635.301. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s Per Watt, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: IS.DSMT OnSMT Off81624324035.7035.4031.2428.07
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNAS Parallel Benchmarks 3.4CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.3 / Avg: 148.47 / Max: 271.57Min: 19.77 / Avg: 150.15 / Max: 264.6Min: 41.72 / Avg: 315.26 / Max: 633.61Min: 40.25 / Avg: 307.61 / Max: 652.71
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: IS.DSMT OnSMT Off2K4K6K8K10KMin: 5245.74 / Avg: 5300.29 / Max: 5419.19Min: 5232.42 / Avg: 5315.15 / Max: 5453.28Min: 8684.84 / Avg: 9849.01 / Max: 10289.96Min: 8012.86 / Avg: 8635.3 / Max: 9435.161. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: BT.CSMT OnSMT Off110K220K330K440K550KSE +/- 396.74, N = 5SE +/- 269.21, N = 5SE +/- 4903.75, N = 15SE +/- 3792.90, N = 12292243.61298801.44491231.83536518.741. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s Per Watt, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: BT.CSMT OnSMT Off300600900120015001328.181382.301236.191408.28
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNAS Parallel Benchmarks 3.4CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 19.79 / Avg: 220.03 / Max: 326.76Min: 19.51 / Avg: 216.16 / Max: 309.45Min: 40.66 / Avg: 397.37 / Max: 663.29Min: 40.15 / Avg: 380.97 / Max: 653.61
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: BT.CSMT OnSMT Off90K180K270K360K450KMin: 291074.28 / Avg: 292243.61 / Max: 293455.75Min: 297934.05 / Avg: 298801.44 / Max: 299587.54Min: 467138.64 / Avg: 491231.83 / Max: 522502.06Min: 504560.39 / Avg: 536518.74 / Max: 547635.31. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: r2c - Backend: FFTW - Precision: float - X Y Z: 512SMT OnSMT Off90180270360450SE +/- 0.52, N = 5SE +/- 0.09, N = 5SE +/- 0.86, N = 7SE +/- 0.77, N = 6245.54248.56433.60430.271. (CXX) g++ options: -O3
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOP/s Per Watt, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: r2c - Backend: FFTW - Precision: float - X Y Z: 512SMT OnSMT Off0.33910.67821.01731.35641.69551.4451.5071.2891.263
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 20.52 / Avg: 169.96 / Max: 279.94Min: 20.14 / Avg: 164.96 / Max: 252.95Min: 42.06 / Avg: 336.38 / Max: 638.26Min: 40.74 / Avg: 340.8 / Max: 617.42
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: r2c - Backend: FFTW - Precision: float - X Y Z: 512SMT OnSMT Off80160240320400Min: 244.04 / Avg: 245.54 / Max: 247Min: 248.29 / Avg: 248.56 / Max: 248.76Min: 431.15 / Avg: 433.6 / Max: 438.45Min: 427.47 / Avg: 430.27 / Max: 432.581. (CXX) g++ options: -O3

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: SP.CSMT OnSMT Off50K100K150K200K250KSE +/- 290.95, N = 4SE +/- 228.86, N = 4SE +/- 1293.24, N = 6SE +/- 1880.70, N = 6131909.91133415.42224243.28231041.571. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s Per Watt, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: SP.CSMT OnSMT Off140280420560700625.95657.40567.82602.82
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNAS Parallel Benchmarks 3.4CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.01 / Avg: 210.74 / Max: 311.07Min: 19.6 / Avg: 202.94 / Max: 285.46Min: 40.07 / Avg: 394.92 / Max: 657.14Min: 41.29 / Avg: 383.27 / Max: 632.69
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: SP.CSMT OnSMT Off40K80K120K160K200KMin: 131089.96 / Avg: 131909.91 / Max: 132464.62Min: 132829.89 / Avg: 133415.42 / Max: 133800.23Min: 220671.08 / Avg: 224243.28 / Max: 228495.38Min: 221923.32 / Avg: 231041.57 / Max: 234228.381. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2

LuxCoreRender

LuxCoreRender is an open-source 3D physically based renderer formerly known as LuxRender. LuxCoreRender supports CPU-based rendering as well as GPU acceleration via OpenCL, NVIDIA CUDA, and NVIDIA OptiX interfaces. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: LuxCore Benchmark - Acceleration: CPUSMT OnSMT Off3691215SE +/- 0.09, N = 15SE +/- 0.08, N = 8SE +/- 0.11, N = 15SE +/- 0.11, N = 1212.188.889.866.97
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgM samples/sec Per Watt, More Is BetterLuxCoreRender 2.6Scene: LuxCore Benchmark - Acceleration: CPUSMT OnSMT Off0.00970.01940.02910.03880.04850.0430.0360.0220.017
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterLuxCoreRender 2.6CPU Power Consumption MonitorSMT OnSMT Off90180270360450Min: 21.47 / Avg: 286.22 / Max: 332.7Min: 20.62 / Avg: 247.27 / Max: 288.16Min: 42.57 / Avg: 445.45 / Max: 519.63Min: 41.4 / Avg: 402.3 / Max: 472.3
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: LuxCore Benchmark - Acceleration: CPUSMT OnSMT Off48121620Min: 11.69 / Avg: 12.18 / Max: 12.8Min: 8.56 / Avg: 8.88 / Max: 9.23Min: 9.24 / Avg: 9.86 / Max: 10.73Min: 6.51 / Avg: 6.97 / Max: 7.87

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: c2c - Backend: FFTW - Precision: float - X Y Z: 512SMT OnSMT Off50100150200250SE +/- 0.15, N = 4SE +/- 0.01, N = 4SE +/- 1.37, N = 5SE +/- 1.28, N = 5128.12128.60221.77223.581. (CXX) g++ options: -O3
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOP/s Per Watt, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: c2c - Backend: FFTW - Precision: float - X Y Z: 512SMT OnSMT Off0.15570.31140.46710.62280.77850.6820.6920.5480.569
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 20.78 / Avg: 187.92 / Max: 271.2Min: 11.69 / Avg: 185.74 / Max: 241.77Min: 44.94 / Avg: 404.61 / Max: 614.02Min: 41.14 / Avg: 392.95 / Max: 599.94
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: c2c - Backend: FFTW - Precision: float - X Y Z: 512SMT OnSMT Off4080120160200Min: 127.73 / Avg: 128.12 / Max: 128.42Min: 128.57 / Avg: 128.6 / Max: 128.63Min: 217.55 / Avg: 221.77 / Max: 224.91Min: 220.34 / Avg: 223.58 / Max: 227.641. (CXX) g++ options: -O3

Primesieve

Primesieve generates prime numbers using a highly optimized sieve of Eratosthenes implementation. Primesieve primarily benchmarks the CPU's L1/L2 cache performance. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterPrimesieve 8.0Length: 1e12SMT OnSMT Off0.43740.87481.31221.74962.187SE +/- 0.006, N = 11SE +/- 0.003, N = 11SE +/- 0.010, N = 14SE +/- 0.003, N = 121.9441.7961.5081.1601. (CXX) g++ options: -O3
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterPrimesieve 8.0CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 20.18 / Avg: 113.25 / Max: 312.47Min: 19.72 / Avg: 104.83 / Max: 309.64Min: 40.74 / Avg: 202.59 / Max: 640.65Min: 40.21 / Avg: 164.82 / Max: 618.35
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterPrimesieve 8.0Length: 1e12SMT OnSMT Off246810Min: 1.92 / Avg: 1.94 / Max: 1.98Min: 1.79 / Avg: 1.8 / Max: 1.82Min: 1.46 / Avg: 1.51 / Max: 1.59Min: 1.15 / Avg: 1.16 / Max: 1.181. (CXX) g++ options: -O3

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Person Vehicle Bike Detection FP16 - Device: CPUSMT OnSMT Off3691215SE +/- 0.08, N = 15SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 310.406.246.466.471. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2022.3CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 21.25 / Avg: 301.97 / Max: 348.43Min: 21.15 / Avg: 267.68 / Max: 294.4Min: 44.24 / Avg: 563.32 / Max: 629.29Min: 199.39 / Avg: 575.68 / Max: 630.57
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Person Vehicle Bike Detection FP16 - Device: CPUSMT OnSMT Off3691215Min: 9.71 / Avg: 10.4 / Max: 10.72Min: 6.23 / Avg: 6.24 / Max: 6.25Min: 6.45 / Avg: 6.46 / Max: 6.47Min: 6.46 / Avg: 6.47 / Max: 6.481. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 256 - Model: GoogLeNetSMT OnSMT Off110220330440550SE +/- 3.39, N = 3SE +/- 0.35, N = 3SE +/- 1.69, N = 3SE +/- 5.52, N = 4504.09525.39329.22452.99
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgimages/sec Per Watt, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 256 - Model: GoogLeNetSMT OnSMT Off0.49860.99721.49581.99442.4932.0722.2160.8591.099
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterTensorFlow 2.12CPU Power Consumption MonitorSMT OnSMT Off80160240320400Min: 20.72 / Avg: 243.24 / Max: 277.92Min: 20.23 / Avg: 237.07 / Max: 270.8Min: 42.83 / Avg: 383.45 / Max: 426.38Min: 189.77 / Avg: 412.23 / Max: 457.74
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 256 - Model: GoogLeNetSMT OnSMT Off90180270360450Min: 497.68 / Avg: 504.09 / Max: 509.2Min: 524.91 / Avg: 525.39 / Max: 526.07Min: 325.97 / Avg: 329.22 / Max: 331.68Min: 445.79 / Avg: 452.99 / Max: 469.44

Appleseed

Appleseed is an open-source production renderer focused on physically-based global illumination rendering engine primarily designed for animation and visual effects. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterAppleseed 2.0 BetaScene: Material TesterSMT OnSMT Off60120180240300166.68167.90265.89
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterAppleseed 2.0 BetaCPU Power Consumption MonitorSMT OnSMT Off50100150200250Min: 21.47 / Avg: 150.04 / Max: 165.18Min: 20.23 / Avg: 148.63 / Max: 164.7Min: 126.26 / Avg: 258.14 / Max: 281.63

Scene: Material Tester

EPYC 9754 2P: SMT On: The test run did not produce a result.

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: FT.CSMT OnSMT Off50K100K150K200K250KSE +/- 1029.98, N = 8SE +/- 93.32, N = 8SE +/- 1757.80, N = 9SE +/- 2978.54, N = 13140791.52147448.72211432.89224178.101. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s Per Watt, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: FT.CSMT OnSMT Off20040060080010001044.171105.80823.02896.92
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNAS Parallel Benchmarks 3.4CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 20.47 / Avg: 134.84 / Max: 312.66Min: 19.73 / Avg: 133.34 / Max: 299.42Min: 40.97 / Avg: 256.9 / Max: 647.79Min: 40.17 / Avg: 249.94 / Max: 637.03
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: FT.CSMT OnSMT Off40K80K120K160K200KMin: 136197.59 / Avg: 140791.52 / Max: 145395.56Min: 147024.3 / Avg: 147448.72 / Max: 147815.8Min: 203018.53 / Avg: 211432.89 / Max: 220067.59Min: 201096.66 / Avg: 224178.1 / Max: 234372.231. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2

7-Zip Compression

This is a test of 7-Zip compression/decompression with its integrated benchmark feature. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 22.01Test: Compression RatingSMT OnSMT Off200K400K600K800K1000KSE +/- 1345.49, N = 3SE +/- 4792.89, N = 3SE +/- 5539.49, N = 3SE +/- 4721.80, N = 37262715927419258207714621. (CXX) g++ options: -lpthread -ldl -O2 -fPIC
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 22.01Test: Compression RatingSMT OnSMT Off160K320K480K640K800KMin: 724114 / Avg: 726271.33 / Max: 728743Min: 583392 / Avg: 592740.67 / Max: 599250Min: 917501 / Avg: 925819.67 / Max: 936316Min: 762298 / Avg: 771461.67 / Max: 7780201. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

OpenVKL

OpenVKL is the Intel Open Volume Kernel Library that offers high-performance volume computation kernels and part of the Intel oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgItems / Sec, More Is BetterOpenVKL 1.3.1Benchmark: vklBenchmark ISPCSMT OnSMT Off400800120016002000SE +/- 1.86, N = 3SE +/- 0.33, N = 3SE +/- 6.06, N = 3SE +/- 1.33, N = 31396110717201530
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgItems / Sec Per Watt, More Is BetterOpenVKL 1.3.1Benchmark: vklBenchmark ISPCSMT OnSMT Off1.31692.63383.95075.26766.58455.8534.9553.6763.512
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenVKL 1.3.1CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 20.44 / Avg: 238.51 / Max: 321.14Min: 19.63 / Avg: 223.41 / Max: 305.39Min: 41.51 / Avg: 467.94 / Max: 643.91Min: 41.74 / Avg: 435.67 / Max: 614.27
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgItems / Sec, More Is BetterOpenVKL 1.3.1Benchmark: vklBenchmark ISPCSMT OnSMT Off30060090012001500Min: 1392 / Avg: 1395.67 / Max: 1398Min: 1107 / Avg: 1107.33 / Max: 1108Min: 1709 / Avg: 1719.67 / Max: 1730Min: 1527 / Avg: 1529.67 / Max: 1531

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 512 - Model: ResNet-50SMT OnSMT Off4080120160200SE +/- 0.90, N = 3SE +/- 1.35, N = 3SE +/- 0.59, N = 3SE +/- 0.13, N = 3122.77123.88172.36189.40
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgimages/sec Per Watt, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 512 - Model: ResNet-50SMT OnSMT Off0.12510.25020.37530.50040.62550.5440.5560.4090.429
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterTensorFlow 2.12CPU Power Consumption MonitorSMT OnSMT Off90180270360450Min: 20.5 / Avg: 225.8 / Max: 286.84Min: 20.41 / Avg: 222.86 / Max: 282.49Min: 42.05 / Avg: 421.61 / Max: 491.54Min: 194 / Avg: 441.9 / Max: 511.48
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 512 - Model: ResNet-50SMT OnSMT Off306090120150Min: 121.87 / Avg: 122.77 / Max: 124.57Min: 122.45 / Avg: 123.88 / Max: 126.57Min: 171.27 / Avg: 172.36 / Max: 173.29Min: 189.2 / Avg: 189.4 / Max: 189.65

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Person Detection FP16 - Device: CPUSMT OnSMT Off918273645SE +/- 0.20, N = 12SE +/- 0.23, N = 9SE +/- 0.14, N = 3SE +/- 0.29, N = 326.6128.7440.6741.051. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Person Detection FP16 - Device: CPUSMT OnSMT Off918273645Min: 26.2 / Avg: 26.61 / Max: 28.75Min: 28.32 / Avg: 28.74 / Max: 30.44Min: 40.42 / Avg: 40.67 / Max: 40.91Min: 40.62 / Avg: 41.05 / Max: 41.611. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Person Detection FP32 - Device: CPUSMT OnSMT Off918273645SE +/- 0.15, N = 15SE +/- 0.29, N = 5SE +/- 0.22, N = 3SE +/- 0.14, N = 326.5728.8340.9040.201. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Person Detection FP32 - Device: CPUSMT OnSMT Off918273645Min: 26.26 / Avg: 26.57 / Max: 28.62Min: 28.41 / Avg: 28.83 / Max: 29.97Min: 40.52 / Avg: 40.9 / Max: 41.28Min: 39.92 / Avg: 40.2 / Max: 40.341. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 512 - Model: GoogLeNetSMT OnSMT Off140280420560700SE +/- 4.74, N = 12SE +/- 5.79, N = 12SE +/- 4.51, N = 15SE +/- 6.73, N = 3416.03429.16538.52634.13
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgimages/sec Per Watt, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 512 - Model: GoogLeNetSMT OnSMT Off0.42280.84561.26841.69122.1141.7671.8791.2271.391
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterTensorFlow 2.12CPU Power Consumption MonitorSMT OnSMT Off90180270360450Min: 20.81 / Avg: 235.45 / Max: 279.59Min: 20.39 / Avg: 228.42 / Max: 276.77Min: 21.61 / Avg: 439.04 / Max: 490.17Min: 192.59 / Avg: 456.01 / Max: 498.3
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 512 - Model: GoogLeNetSMT OnSMT Off110220330440550Min: 408.57 / Avg: 416.03 / Max: 467.84Min: 419.48 / Avg: 429.16 / Max: 492.59Min: 516.28 / Avg: 538.52 / Max: 558.56Min: 620.78 / Avg: 634.13 / Max: 642.24

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Age Gender Recognition Retail 0013 FP16 - Device: CPUSMT OnSMT Off40K80K120K160K200KSE +/- 390.62, N = 3SE +/- 202.90, N = 3SE +/- 676.62, N = 3SE +/- 158.47, N = 3120515.11113162.29168372.89142135.581. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Age Gender Recognition Retail 0013 FP16 - Device: CPUSMT OnSMT Off30K60K90K120K150KMin: 119875.09 / Avg: 120515.11 / Max: 121223.11Min: 112916.46 / Avg: 113162.29 / Max: 113564.82Min: 167350.19 / Avg: 168372.89 / Max: 169651.7Min: 141829.04 / Avg: 142135.58 / Max: 142358.571. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: CG.CSMT OnSMT Off14K28K42K56K70KSE +/- 441.00, N = 15SE +/- 285.31, N = 8SE +/- 348.18, N = 8SE +/- 568.39, N = 845686.8848672.2467554.7466822.971. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s Per Watt, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: CG.CSMT OnSMT Off70140210280350304.95340.22246.35257.05
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNAS Parallel Benchmarks 3.4CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 20.14 / Avg: 149.82 / Max: 321.08Min: 19.7 / Avg: 143.06 / Max: 313.46Min: 41.39 / Avg: 274.22 / Max: 646.25Min: 40.04 / Avg: 259.96 / Max: 632.64
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: CG.CSMT OnSMT Off12K24K36K48K60KMin: 42915.05 / Avg: 45686.88 / Max: 47917.08Min: 46884.47 / Avg: 48672.24 / Max: 49512.56Min: 65952.73 / Avg: 67554.74 / Max: 68895.94Min: 63255.37 / Avg: 66822.97 / Max: 68389.791. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2

Timed Linux Kernel Compilation

This test times how long it takes to build the Linux kernel in a default configuration (defconfig) for the architecture being tested or alternatively an allmodconfig for building all possible kernel modules for the build. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.1Build: defconfigSMT OnSMT Off612182430SE +/- 0.23, N = 7SE +/- 0.21, N = 7SE +/- 0.14, N = 13SE +/- 0.12, N = 1326.2322.9820.3418.47
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterTimed Linux Kernel Compilation 6.1CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.11 / Avg: 158.04 / Max: 312.79Min: 19.82 / Avg: 149.41 / Max: 323.82Min: 41.05 / Avg: 285.97 / Max: 638.63Min: 40.17 / Avg: 267.88 / Max: 656.25
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.1Build: defconfigSMT OnSMT Off612182430Min: 25.84 / Avg: 26.23 / Max: 27.6Min: 22.72 / Avg: 22.98 / Max: 24.26Min: 20.05 / Avg: 20.34 / Max: 21.94Min: 18.3 / Avg: 18.47 / Max: 19.94

LuxCoreRender

LuxCoreRender is an open-source 3D physically based renderer formerly known as LuxRender. LuxCoreRender supports CPU-based rendering as well as GPU acceleration via OpenCL, NVIDIA CUDA, and NVIDIA OptiX interfaces. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: DLSC - Acceleration: CPUSMT OnSMT Off510152025SE +/- 0.10, N = 3SE +/- 0.05, N = 3SE +/- 0.20, N = 4SE +/- 0.10, N = 316.3413.2718.6114.45
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgM samples/sec Per Watt, More Is BetterLuxCoreRender 2.6Scene: DLSC - Acceleration: CPUSMT OnSMT Off0.01220.02440.03660.04880.0610.0540.0470.0360.031
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterLuxCoreRender 2.6CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.94 / Avg: 303.49 / Max: 332.56Min: 20.7 / Avg: 285.26 / Max: 319.31Min: 41.78 / Avg: 510.83 / Max: 659.11Min: 41.49 / Avg: 468.61 / Max: 635.46
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: DLSC - Acceleration: CPUSMT OnSMT Off510152025Min: 16.14 / Avg: 16.34 / Max: 16.49Min: 13.16 / Avg: 13.27 / Max: 13.34Min: 18.33 / Avg: 18.61 / Max: 19.18Min: 14.32 / Avg: 14.45 / Max: 14.66

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-StreamSMT OnSMT Off2004006008001000SE +/- 0.14, N = 3SE +/- 0.16, N = 3SE +/- 0.27, N = 3SE +/- 0.02, N = 3859.87644.98902.86676.81
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.51 / Avg: 247.92 / Max: 356.8Min: 20.07 / Avg: 233.36 / Max: 344.64Min: 43.81 / Avg: 482.63 / Max: 702.05Min: 194.94 / Avg: 478.36 / Max: 692.61
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-StreamSMT OnSMT Off160320480640800Min: 859.64 / Avg: 859.87 / Max: 860.11Min: 644.76 / Avg: 644.98 / Max: 645.29Min: 902.36 / Avg: 902.86 / Max: 903.28Min: 676.77 / Avg: 676.81 / Max: 676.85

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-StreamSMT OnSMT Off2004006008001000SE +/- 0.49, N = 3SE +/- 0.12, N = 3SE +/- 0.45, N = 3SE +/- 0.21, N = 3859.88645.00902.78676.43
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.1 / Avg: 250.83 / Max: 357.61Min: 20.77 / Avg: 235.91 / Max: 344.85Min: 44.26 / Avg: 486.69 / Max: 702.22Min: 192.79 / Avg: 479.5 / Max: 693.71
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-StreamSMT OnSMT Off160320480640800Min: 859.29 / Avg: 859.88 / Max: 860.84Min: 644.87 / Avg: 645 / Max: 645.24Min: 901.88 / Avg: 902.78 / Max: 903.34Min: 676.02 / Avg: 676.43 / Max: 676.71

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-StreamSMT OnSMT Off1530456075SE +/- 0.05, N = 3SE +/- 0.20, N = 3SE +/- 0.06, N = 3SE +/- 0.04, N = 365.9749.5168.3651.81
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.26 / Avg: 279.75 / Max: 331.26Min: 20.35 / Avg: 278.52 / Max: 329.31Min: 43.24 / Avg: 556.63 / Max: 669.25Min: 193.12 / Avg: 553.79 / Max: 663.52
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-StreamSMT OnSMT Off1326395265Min: 65.89 / Avg: 65.97 / Max: 66.05Min: 49.11 / Avg: 49.51 / Max: 49.74Min: 68.27 / Avg: 68.36 / Max: 68.48Min: 51.75 / Avg: 51.81 / Max: 51.89

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-StreamSMT OnSMT Off20406080100SE +/- 0.04, N = 3SE +/- 0.37, N = 3SE +/- 0.07, N = 3SE +/- 0.05, N = 3102.2377.81107.1881.20
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.93 / Avg: 256.18 / Max: 318.33Min: 20.87 / Avg: 257.43 / Max: 318.21Min: 43.67 / Avg: 535.06 / Max: 668.18Min: 196.83 / Avg: 534.41 / Max: 647.28
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-StreamSMT OnSMT Off20406080100Min: 102.15 / Avg: 102.23 / Max: 102.3Min: 77.21 / Avg: 77.81 / Max: 78.48Min: 107.07 / Avg: 107.18 / Max: 107.31Min: 81.11 / Avg: 81.2 / Max: 81.28

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-StreamSMT OnSMT Off1122334455SE +/- 0.04, N = 3SE +/- 0.10, N = 3SE +/- 0.13, N = 3SE +/- 0.57, N = 1546.4335.5548.6245.88
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.92 / Avg: 250.7 / Max: 320.58Min: 20.36 / Avg: 251.27 / Max: 319.12Min: 42.64 / Avg: 517.56 / Max: 680.99Min: 193.54 / Avg: 505.33 / Max: 653.26
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-StreamSMT OnSMT Off1020304050Min: 46.35 / Avg: 46.43 / Max: 46.48Min: 35.38 / Avg: 35.55 / Max: 35.72Min: 48.39 / Avg: 48.62 / Max: 48.82Min: 42.17 / Avg: 45.88 / Max: 50.1

Intel Open Image Denoise

Open Image Denoise is a denoising library for ray-tracing and part of the Intel oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 2.0Run: RTLightmap.hdr.4096x4096 - Device: CPU-OnlySMT OnSMT Off0.52881.05761.58642.11522.644SE +/- 0.00, N = 4SE +/- 0.01, N = 4SE +/- 0.00, N = 5SE +/- 0.02, N = 51.741.722.352.09
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgImages / Sec Per Watt, More Is BetterIntel Open Image Denoise 2.0Run: RTLightmap.hdr.4096x4096 - Device: CPU-OnlySMT OnSMT Off0.00180.00360.00540.00720.0090.0080.0080.0060.006
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterIntel Open Image Denoise 2.0CPU Power Consumption MonitorSMT OnSMT Off100200300400500Min: 20.37 / Avg: 219.82 / Max: 297.83Min: 19.75 / Avg: 215.85 / Max: 296.82Min: 41.62 / Avg: 383.91 / Max: 567.02Min: 41.02 / Avg: 379.66 / Max: 554.87
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 2.0Run: RTLightmap.hdr.4096x4096 - Device: CPU-OnlySMT OnSMT Off246810Min: 1.74 / Avg: 1.74 / Max: 1.74Min: 1.71 / Avg: 1.72 / Max: 1.75Min: 2.34 / Avg: 2.35 / Max: 2.36Min: 2.04 / Avg: 2.09 / Max: 2.14

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-StreamSMT OnSMT Off50100150200250SE +/- 0.54, N = 3SE +/- 0.81, N = 3SE +/- 0.18, N = 3SE +/- 0.09, N = 3201.63156.06211.48162.02
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption MonitorSMT OnSMT Off130260390520650Min: 20.88 / Avg: 247.32 / Max: 358.25Min: 20.7 / Avg: 235.42 / Max: 345.62Min: 41.6 / Avg: 459.55 / Max: 712.43Min: 195.28 / Avg: 482.32 / Max: 698.28
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-StreamSMT OnSMT Off4080120160200Min: 200.56 / Avg: 201.63 / Max: 202.21Min: 154.6 / Avg: 156.06 / Max: 157.38Min: 211.27 / Avg: 211.48 / Max: 211.84Min: 161.84 / Avg: 162.02 / Max: 162.15

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-StreamSMT OnSMT Off4080120160200SE +/- 0.16, N = 3SE +/- 1.18, N = 15SE +/- 0.05, N = 3SE +/- 0.10, N = 3152.99125.45159.89118.13
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.8 / Avg: 291.22 / Max: 348.67Min: 20.34 / Avg: 281.81 / Max: 336.34Min: 43.03 / Avg: 568.51 / Max: 686.75Min: 194.59 / Avg: 576 / Max: 679.21
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-StreamSMT OnSMT Off306090120150Min: 152.74 / Avg: 152.99 / Max: 153.28Min: 112.71 / Avg: 125.45 / Max: 128.15Min: 159.79 / Avg: 159.89 / Max: 159.98Min: 117.95 / Avg: 118.13 / Max: 118.29

MariaDB

This is a MariaDB MySQL database server benchmark making use of mysqlslap. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.0.1Clients: 2048SMT OnSMT Off2004006008001000SE +/- 1.81, N = 3SE +/- 1.06, N = 3SE +/- 8.04, N = 3SE +/- 0.73, N = 37807835805911. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lnuma -lpcre2-8 -lcrypt -lz -lm -lssl -lcrypto -lpthread -ldl
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgQueries Per Second Per Watt, More Is BetterMariaDB 11.0.1Clients: 2048SMT OnSMT Off2468106.3716.3632.3652.322
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterMariaDB 11.0.1CPU Power Consumption MonitorSMT OnSMT Off50100150200250Min: 20.42 / Avg: 122.43 / Max: 135.29Min: 20.3 / Avg: 123.06 / Max: 134.69Min: 43.1 / Avg: 245.22 / Max: 266.18Min: 188.47 / Avg: 254.51 / Max: 270.43
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.0.1Clients: 2048SMT OnSMT Off140280420560700Min: 777.6 / Avg: 780.24 / Max: 783.7Min: 781.25 / Avg: 783.09 / Max: 784.93Min: 565.61 / Avg: 579.71 / Max: 593.47Min: 589.28 / Avg: 590.67 / Max: 591.721. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lnuma -lpcre2-8 -lcrypt -lz -lm -lssl -lcrypto -lpthread -ldl

Appleseed

Appleseed is an open-source production renderer focused on physically-based global illumination rendering engine primarily designed for animation and visual effects. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterAppleseed 2.0 BetaScene: EmilySMT OnSMT Off4080120160200122.59123.30164.57159.91
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterAppleseed 2.0 BetaCPU Power Consumption MonitorSMT OnSMT Off60120180240300Min: 21.17 / Avg: 177.28 / Max: 240.78Min: 20.77 / Avg: 173.02 / Max: 236.67Min: 44.3 / Avg: 290.19 / Max: 351.48Min: 199.89 / Avg: 289.71 / Max: 346.75

Intel Open Image Denoise

Open Image Denoise is a denoising library for ray-tracing and part of the Intel oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 2.0Run: RT.ldr_alb_nrm.3840x2160 - Device: CPU-OnlySMT OnSMT Off1.07552.1513.22654.3025.3775SE +/- 0.00, N = 7SE +/- 0.02, N = 15SE +/- 0.01, N = 7SE +/- 0.03, N = 153.623.574.784.31
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgImages / Sec Per Watt, More Is BetterIntel Open Image Denoise 2.0Run: RT.ldr_alb_nrm.3840x2160 - Device: CPU-OnlySMT OnSMT Off0.00470.00940.01410.01880.02350.0210.0210.0160.015
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterIntel Open Image Denoise 2.0CPU Power Consumption MonitorSMT OnSMT Off100200300400500Min: 20.39 / Avg: 170.54 / Max: 297.28Min: 19.82 / Avg: 174.13 / Max: 300.84Min: 40.73 / Avg: 294.58 / Max: 558.39Min: 40.95 / Avg: 295.21 / Max: 552.75
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 2.0Run: RT.ldr_alb_nrm.3840x2160 - Device: CPU-OnlySMT OnSMT Off246810Min: 3.61 / Avg: 3.62 / Max: 3.63Min: 3.38 / Avg: 3.57 / Max: 3.69Min: 4.74 / Avg: 4.78 / Max: 4.82Min: 4.04 / Avg: 4.31 / Max: 4.51

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Age Gender Recognition Retail 0013 FP16 - Device: CPUSMT OnSMT Off0.18680.37360.56040.74720.934SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.830.700.620.671. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2022.3CPU Power Consumption MonitorSMT OnSMT Off100200300400500Min: 21.1 / Avg: 320.15 / Max: 350.37Min: 21.12 / Avg: 294.6 / Max: 327.84Min: 42.63 / Avg: 504.47 / Max: 594.16Min: 199.33 / Avg: 511.7 / Max: 565.18
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Age Gender Recognition Retail 0013 FP16 - Device: CPUSMT OnSMT Off246810Min: 0.83 / Avg: 0.83 / Max: 0.84Min: 0.69 / Avg: 0.7 / Max: 0.7Min: 0.62 / Avg: 0.62 / Max: 0.62Min: 0.66 / Avg: 0.67 / Max: 0.671. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

Aircrack-ng

Aircrack-ng is a tool for assessing WiFi/WLAN network security. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgk/s, More Is BetterAircrack-ng 1.7SMT OnSMT Off40K80K120K160K200KSE +/- 101.44, N = 3SE +/- 1020.90, N = 3SE +/- 1109.71, N = 3171120.35149056.95128050.221. (CXX) g++ options: -std=gnu++17 -O3 -fvisibility=hidden -fcommon -rdynamic -lnl-3 -lnl-genl-3 -lpcre -lpthread -lz -lssl -lcrypto -lhwloc -ldl -lm -pthread
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgk/s Per Watt, More Is BetterAircrack-ng 1.7SMT OnSMT Off160320480640800729.35649.61352.95
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterAircrack-ng 1.7CPU Power Consumption MonitorSMT OnSMT Off70140210280350Min: 21.31 / Avg: 234.62 / Max: 267.15Min: 20.61 / Avg: 229.46 / Max: 260.41Min: 42.14 / Avg: 362.8 / Max: 415.41
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgk/s, More Is BetterAircrack-ng 1.7SMT OnSMT Off30K60K90K120K150KMin: 170978.7 / Avg: 171120.35 / Max: 171316.97Min: 147034.39 / Avg: 149056.95 / Max: 150310.39Min: 126334.87 / Avg: 128050.22 / Max: 130127.561. (CXX) g++ options: -std=gnu++17 -O3 -fvisibility=hidden -fcommon -rdynamic -lnl-3 -lnl-genl-3 -lpcre -lpthread -lz -lssl -lcrypto -lhwloc -ldl -lm -pthread

EPYC 9754 2P: SMT On: The test run did not produce a result.

Intel Open Image Denoise

Open Image Denoise is a denoising library for ray-tracing and part of the Intel oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 2.0Run: RT.hdr_alb_nrm.3840x2160 - Device: CPU-OnlySMT OnSMT Off1.0712.1423.2134.2845.355SE +/- 0.00, N = 7SE +/- 0.02, N = 7SE +/- 0.01, N = 7SE +/- 0.03, N = 73.623.574.764.35
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgImages / Sec Per Watt, More Is BetterIntel Open Image Denoise 2.0Run: RT.hdr_alb_nrm.3840x2160 - Device: CPU-OnlySMT OnSMT Off0.00470.00940.01410.01880.02350.0210.0210.0160.015
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterIntel Open Image Denoise 2.0CPU Power Consumption MonitorSMT OnSMT Off100200300400500Min: 20.3 / Avg: 176.53 / Max: 298.39Min: 19.89 / Avg: 172.35 / Max: 299.67Min: 41.67 / Avg: 298.49 / Max: 561.86Min: 21.72 / Avg: 294.18 / Max: 551.63
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 2.0Run: RT.hdr_alb_nrm.3840x2160 - Device: CPU-OnlySMT OnSMT Off246810Min: 3.61 / Avg: 3.62 / Max: 3.62Min: 3.52 / Avg: 3.57 / Max: 3.67Min: 4.72 / Avg: 4.76 / Max: 4.78Min: 4.25 / Avg: 4.35 / Max: 4.48

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 256 - Model: AlexNetSMT OnSMT Off30060090012001500SE +/- 3.81, N = 3SE +/- 8.29, N = 3SE +/- 14.23, N = 15SE +/- 11.49, N = 151422.081375.331225.411581.66
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgimages/sec Per Watt, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 256 - Model: AlexNetSMT OnSMT Off2468107.0587.1773.6654.206
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterTensorFlow 2.12CPU Power Consumption MonitorSMT OnSMT Off80160240320400Min: 20.26 / Avg: 201.49 / Max: 263.75Min: 19.84 / Avg: 191.63 / Max: 248.98Min: 41.4 / Avg: 334.38 / Max: 420.36Min: 194.45 / Avg: 376.07 / Max: 452.14
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 256 - Model: AlexNetSMT OnSMT Off30060090012001500Min: 1414.88 / Avg: 1422.08 / Max: 1427.83Min: 1360.09 / Avg: 1375.33 / Max: 1388.62Min: 1138.49 / Avg: 1225.41 / Max: 1316.61Min: 1499.67 / Avg: 1581.66 / Max: 1638.97

MariaDB

This is a MariaDB MySQL database server benchmark making use of mysqlslap. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.0.1Clients: 4096SMT OnSMT Off150300450600750SE +/- 8.08, N = 3SE +/- 3.99, N = 3SE +/- 5.45, N = 6SE +/- 1.48, N = 36556955455791. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lnuma -lpcre2-8 -lcrypt -lz -lm -lssl -lcrypto -lpthread -ldl
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgQueries Per Second Per Watt, More Is BetterMariaDB 11.0.1Clients: 4096SMT OnSMT Off1.26952.5393.80855.0786.34755.3675.6422.1762.249
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterMariaDB 11.0.1CPU Power Consumption MonitorSMT OnSMT Off50100150200250Min: 20.08 / Avg: 122.04 / Max: 137.54Min: 19.33 / Avg: 123.19 / Max: 137.52Min: 40.82 / Avg: 250.44 / Max: 275.37Min: 188.93 / Avg: 257.46 / Max: 279.59
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.0.1Clients: 4096SMT OnSMT Off120240360480600Min: 641.44 / Avg: 654.79 / Max: 669.34Min: 690.61 / Avg: 695.29 / Max: 703.23Min: 523.29 / Avg: 545.33 / Max: 559.91Min: 576.04 / Avg: 578.82 / Max: 581.061. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lnuma -lpcre2-8 -lcrypt -lz -lm -lssl -lcrypto -lpthread -ldl

Timed LLVM Compilation

This test times how long it takes to compile/build the LLVM compiler stack. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 16.0Build System: NinjaSMT OnSMT Off306090120150SE +/- 0.29, N = 3SE +/- 0.18, N = 3SE +/- 0.91, N = 3SE +/- 0.70, N = 3125.36119.58107.7299.22
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterTimed LLVM Compilation 16.0CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.62 / Avg: 209.56 / Max: 335.43Min: 20.32 / Avg: 218.43 / Max: 329.81Min: 42.36 / Avg: 337.35 / Max: 671.56Min: 41.09 / Avg: 351.2 / Max: 660.92
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 16.0Build System: NinjaSMT OnSMT Off20406080100Min: 124.8 / Avg: 125.36 / Max: 125.73Min: 119.38 / Avg: 119.58 / Max: 119.93Min: 106.42 / Avg: 107.72 / Max: 109.47Min: 98.22 / Avg: 99.21 / Max: 100.58

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 512 - Model: AlexNetSMT OnSMT Off400800120016002000SE +/- 1.79, N = 3SE +/- 3.13, N = 3SE +/- 15.22, N = 15SE +/- 18.22, N = 31628.801526.811770.911908.76
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgimages/sec Per Watt, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 512 - Model: AlexNetSMT OnSMT Off2468106.9227.1594.5844.540
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterTensorFlow 2.12CPU Power Consumption MonitorSMT OnSMT Off90180270360450Min: 20.29 / Avg: 235.29 / Max: 286.59Min: 19.92 / Avg: 213.28 / Max: 261.74Min: 42.41 / Avg: 386.28 / Max: 479.62Min: 192.51 / Avg: 420.4 / Max: 490.99
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 512 - Model: AlexNetSMT OnSMT Off30060090012001500Min: 1625.25 / Avg: 1628.8 / Max: 1631.02Min: 1520.65 / Avg: 1526.81 / Max: 1530.86Min: 1670.92 / Avg: 1770.91 / Max: 1845.71Min: 1881.06 / Avg: 1908.76 / Max: 1943.1

Timed Node.js Compilation

This test profile times how long it takes to build/compile Node.js itself from source. Node.js is a JavaScript run-time built from the Chrome V8 JavaScript engine while itself is written in C/C++. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterTimed Node.js Compilation 19.8.1Time To CompileSMT OnSMT Off306090120150SE +/- 0.01, N = 3SE +/- 0.07, N = 3SE +/- 0.09, N = 3SE +/- 0.04, N = 3113.70116.1293.2793.95
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterTimed Node.js Compilation 19.8.1CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.49 / Avg: 192.96 / Max: 340.93Min: 19.79 / Avg: 195.41 / Max: 330.02Min: 40.27 / Avg: 330.87 / Max: 680.38Min: 40.21 / Avg: 335.02 / Max: 662.48
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterTimed Node.js Compilation 19.8.1Time To CompileSMT OnSMT Off20406080100Min: 113.69 / Avg: 113.7 / Max: 113.72Min: 115.99 / Avg: 116.12 / Max: 116.22Min: 93.16 / Avg: 93.27 / Max: 93.45Min: 93.89 / Avg: 93.95 / Max: 94.04

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 256 - Model: ResNet-50SMT OnSMT Off306090120150SE +/- 1.07, N = 12SE +/- 1.45, N = 12SE +/- 1.70, N = 3SE +/- 0.68, N = 3118.45121.47124.98146.74
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgimages/sec Per Watt, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 256 - Model: ResNet-50SMT OnSMT Off0.12290.24580.36870.49160.61450.5360.5460.3190.361
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterTensorFlow 2.12CPU Power Consumption MonitorSMT OnSMT Off80160240320400Min: 20.64 / Avg: 220.96 / Max: 274.28Min: 20.24 / Avg: 222.59 / Max: 275.13Min: 42.84 / Avg: 391.59 / Max: 458.74Min: 194.78 / Avg: 406.75 / Max: 462.03
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 256 - Model: ResNet-50SMT OnSMT Off306090120150Min: 116.73 / Avg: 118.45 / Max: 130.19Min: 119.56 / Avg: 121.47 / Max: 137.45Min: 122.09 / Avg: 124.98 / Max: 127.96Min: 145.65 / Avg: 146.74 / Max: 147.98

miniFE

MiniFE Finite Element is an application for unstructured implicit finite element codes. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgCG Mflops, More Is BetterminiFE 2.2Problem Size: SmallSMT OnSMT Off13K26K39K52K65KSE +/- 52.11, N = 5SE +/- 25.22, N = 5SE +/- 411.42, N = 5SE +/- 478.13, N = 551784.151741.062774.253798.61. (CXX) g++ options: -O3 -fopenmp -lmpi_cxx -lmpi
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgCG Mflops Per Watt, More Is BetterminiFE 2.2Problem Size: SmallSMT OnSMT Off90180270360450419.54430.07268.29236.18
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterminiFE 2.2CPU Power Consumption MonitorSMT OnSMT Off80160240320400Min: 20.05 / Avg: 123.43 / Max: 252.15Min: 19.39 / Avg: 120.31 / Max: 246.66Min: 39.79 / Avg: 233.98 / Max: 478.7Min: 39.96 / Avg: 227.79 / Max: 468.68
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgCG Mflops, More Is BetterminiFE 2.2Problem Size: SmallSMT OnSMT Off11K22K33K44K55KMin: 51648.6 / Avg: 51784.12 / Max: 51952.8Min: 51643.6 / Avg: 51740.98 / Max: 51779Min: 61659 / Avg: 62774.2 / Max: 63844Min: 52456.6 / Avg: 53798.58 / Max: 55417.11. (CXX) g++ options: -O3 -fopenmp -lmpi_cxx -lmpi

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPUSMT OnSMT Off0.29250.5850.87751.171.4625SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 31.211.301.081.131. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2022.3CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.48 / Avg: 310.01 / Max: 336.95Min: 20.96 / Avg: 278.47 / Max: 305.64Min: 42.85 / Avg: 599.24 / Max: 664.03Min: 196.32 / Avg: 577.61 / Max: 653.4
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPUSMT OnSMT Off246810Min: 1.21 / Avg: 1.21 / Max: 1.22Min: 1.29 / Avg: 1.3 / Max: 1.3Min: 1.08 / Avg: 1.08 / Max: 1.08Min: 1.12 / Avg: 1.13 / Max: 1.131. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

Appleseed

Appleseed is an open-source production renderer focused on physically-based global illumination rendering engine primarily designed for animation and visual effects. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterAppleseed 2.0 BetaScene: Disney MaterialSMT OnSMT Off102030405044.3038.4940.57
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterAppleseed 2.0 BetaCPU Power Consumption MonitorSMT OnSMT Off70140210280350Min: 20.47 / Avg: 219.92 / Max: 241.15Min: 20.19 / Avg: 222.18 / Max: 246.09Min: 193.93 / Avg: 357.2 / Max: 399.57

Scene: Disney Material

EPYC 9754 2P: SMT On: The test run did not produce a result.

Timed Gem5 Compilation

This test times how long it takes to compile Gem5. Gem5 is a simulator for computer system architecture research. Gem5 is widely used for computer architecture research within the industry, academia, and more. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterTimed Gem5 Compilation 21.2Time To CompileSMT OnSMT Off4080120160200SE +/- 0.32, N = 3SE +/- 0.50, N = 3SE +/- 1.29, N = 3SE +/- 1.24, N = 3161.65148.48152.59148.38
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterTimed Gem5 Compilation 21.2CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 19.75 / Avg: 126.31 / Max: 329.48Min: 19.52 / Avg: 131.21 / Max: 323.7Min: 43.13 / Avg: 253.79 / Max: 657.39Min: 40.34 / Avg: 250.05 / Max: 652.68
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterTimed Gem5 Compilation 21.2Time To CompileSMT OnSMT Off306090120150Min: 161.05 / Avg: 161.65 / Max: 162.17Min: 147.5 / Avg: 148.48 / Max: 149.05Min: 150.03 / Avg: 152.59 / Max: 154.15Min: 145.89 / Avg: 148.38 / Max: 149.7

Timed LLVM Compilation

This test times how long it takes to compile/build the LLVM compiler stack. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 16.0Build System: Unix MakefilesSMT OnSMT Off50100150200250SE +/- 0.15, N = 3SE +/- 0.70, N = 3SE +/- 0.14, N = 3SE +/- 0.11, N = 3211.08213.91199.37198.75
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterTimed LLVM Compilation 16.0CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 20.39 / Avg: 159.32 / Max: 327.5Min: 19.66 / Avg: 159.92 / Max: 323.96Min: 41.07 / Avg: 282.76 / Max: 651.4Min: 41.34 / Avg: 283.34 / Max: 646.29
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 16.0Build System: Unix MakefilesSMT OnSMT Off4080120160200Min: 210.81 / Avg: 211.08 / Max: 211.32Min: 212.7 / Avg: 213.91 / Max: 215.11Min: 199.09 / Avg: 199.36 / Max: 199.52Min: 198.62 / Avg: 198.75 / Max: 198.98

Timed Godot Game Engine Compilation

This test times how long it takes to compile the Godot Game Engine. Godot is a popular, open-source, cross-platform 2D/3D game engine and is built using the SCons build system and targeting the X11 platform. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterTimed Godot Game Engine Compilation 4.0Time To CompileSMT OnSMT Off20406080100SE +/- 0.14, N = 3SE +/- 0.17, N = 3SE +/- 1.01, N = 3SE +/- 0.30, N = 3105.80102.59100.79100.24
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterTimed Godot Game Engine Compilation 4.0CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.04 / Avg: 148.45 / Max: 334.92Min: 19.83 / Avg: 154.5 / Max: 328.65Min: 41.9 / Avg: 275.11 / Max: 667.57Min: 40.04 / Avg: 275.62 / Max: 665.8
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterTimed Godot Game Engine Compilation 4.0Time To CompileSMT OnSMT Off20406080100Min: 105.53 / Avg: 105.8 / Max: 105.96Min: 102.34 / Avg: 102.59 / Max: 102.92Min: 99.29 / Avg: 100.79 / Max: 102.71Min: 99.66 / Avg: 100.24 / Max: 100.67

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Weld Porosity Detection FP16-INT8 - Device: CPUSMT OnSMT Off3691215SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 310.8410.9111.0011.101. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2022.3CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.98 / Avg: 307.35 / Max: 332.52Min: 20.67 / Avg: 301.27 / Max: 325.39Min: 43.92 / Avg: 603.75 / Max: 664.52Min: 198.25 / Avg: 614.16 / Max: 663.84
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Weld Porosity Detection FP16-INT8 - Device: CPUSMT OnSMT Off3691215Min: 10.84 / Avg: 10.84 / Max: 10.84Min: 10.88 / Avg: 10.91 / Max: 10.92Min: 11 / Avg: 11 / Max: 11.01Min: 11.09 / Avg: 11.1 / Max: 11.111. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

nekRS

nekRS is an open-source Navier Stokes solver based on the spectral element method. NekRS supports both CPU and GPU/accelerator support though this test profile is currently configured for CPU execution. NekRS is part of Nek5000 of the Mathematics and Computer Science MCS at Argonne National Laboratory. This nekRS benchmark is primarily relevant to large core count HPC servers and otherwise may be very time consuming on smaller systems. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgflops/rank, More Is BetternekRS 23.0Input: KershawSMT OnSMT Off1200M2400M3600M4800M6000MSE +/- 8226739.60, N = 3SE +/- 37190783.06, N = 3580863666757342666671. (CXX) g++ options: -fopenmp -O2 -march=native -mtune=native -ftree-vectorize -rdynamic -lmpi_cxx -lmpi
OpenBenchmarking.orgflops/rank Per Watt, More Is BetternekRS 23.0Input: KershawSMT OnSMT Off4M8M12M16M20M19786057.9919236505.51
OpenBenchmarking.orgWatts, Fewer Is BetternekRS 23.0CPU Power Consumption MonitorSMT OnSMT Off60120180240300Min: 21.32 / Avg: 293.57 / Max: 349.54Min: 20.08 / Avg: 298.09 / Max: 347.49
OpenBenchmarking.orgflops/rank, More Is BetternekRS 23.0Input: KershawSMT OnSMT Off1000M2000M3000M4000M5000MMin: 5792190000 / Avg: 5808636666.67 / Max: 5817270000Min: 5667210000 / Avg: 5734266666.67 / Max: 57956700001. (CXX) g++ options: -fopenmp -O2 -march=native -mtune=native -ftree-vectorize -rdynamic -lmpi_cxx -lmpi

Input: Kershaw

EPYC 9754 2P: SMT On: The test quit with a non-zero exit status. E: mpirun noticed that process rank 0 with PID 0 on node phoronix-QuantaGrid-D54Q-2U exited on signal 11 (Segmentation fault).

EPYC 9754 2P: SMT Off: The test quit with a non-zero exit status. E: mpirun noticed that process rank 0 with PID 0 on node phoronix-QuantaGrid-D54Q-2U exited on signal 11 (Segmentation fault).

OpenBenchmarking.orgWatts, Fewer Is BetternekRS 23.0CPU Power Consumption MonitorSMT OnSMT Off60120180240300Min: 21.11 / Avg: 288.26 / Max: 349.64Min: 20.65 / Avg: 287.58 / Max: 350.2

OpenBenchmarking.orgflops/rank Per Watt, More Is BetternekRS 23.0Input: TurboPipe PeriodicSMT OnSMT Off2M4M6M8M10M8806106.198992432.84

CPU Power Consumption Monitor

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWattsCPU Power Consumption MonitorPhoronix Test Suite System MonitoringSMT OnSMT Off140280420560700Min: 10.52 / Avg: 248.93 / Max: 397.25Min: 10.61 / Avg: 238.75 / Max: 362.1Min: 21.61 / Avg: 460.57 / Max: 792.38Min: 97.83 / Avg: 446.01 / Max: 702.85

OpenVINO

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2022.3CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.17 / Avg: 297.12 / Max: 335.54Min: 20.85 / Avg: 299.11 / Max: 323.89Min: 42.91 / Avg: 602.26 / Max: 657.52Min: 195.61 / Avg: 613.08 / Max: 656.04

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Vehicle Detection FP16-INT8 - Device: CPUSMT OnSMT Off3691215SE +/- 0.20, N = 15SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 312.074.744.834.861. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2022.3CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.17 / Avg: 297.12 / Max: 335.54Min: 20.85 / Avg: 299.11 / Max: 323.89Min: 42.91 / Avg: 602.26 / Max: 657.52Min: 195.61 / Avg: 613.08 / Max: 656.04
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Vehicle Detection FP16-INT8 - Device: CPUSMT OnSMT Off48121620Min: 9.59 / Avg: 12.07 / Max: 12.6Min: 4.73 / Avg: 4.74 / Max: 4.74Min: 4.83 / Avg: 4.83 / Max: 4.83Min: 4.85 / Avg: 4.86 / Max: 4.861. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Vehicle Detection FP16-INT8 - Device: CPUSMT OnSMT Off3K6K9K12K15KSE +/- 103.73, N = 15SE +/- 3.11, N = 3SE +/- 7.09, N = 3SE +/- 2.74, N = 35311.486744.6613214.2013141.511. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Vehicle Detection FP16-INT8 - Device: CPUSMT OnSMT Off2K4K6K8K10KMin: 5064.62 / Avg: 5311.48 / Max: 6661.38Min: 6740.02 / Avg: 6744.66 / Max: 6750.56Min: 13205.44 / Avg: 13214.2 / Max: 13228.25Min: 13138.12 / Avg: 13141.51 / Max: 13146.931. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Vehicle Detection FP16 - Device: CPUSMT OnSMT Off13002600390052006500SE +/- 21.50, N = 14SE +/- 25.69, N = 13SE +/- 93.68, N = 15SE +/- 87.31, N = 131464.292784.675810.416242.681. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Vehicle Detection FP16 - Device: CPUSMT OnSMT Off11002200330044005500Min: 1419.29 / Avg: 1464.29 / Max: 1738.62Min: 2724.09 / Avg: 2784.67 / Max: 3086.52Min: 5617.04 / Avg: 5810.41 / Max: 7099.08Min: 6061.1 / Avg: 6242.68 / Max: 7280.181. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF

Neural Magic DeepSparse

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.09 / Avg: 235.18 / Max: 327.39Min: 20.67 / Avg: 228.97 / Max: 321.66Min: 42.89 / Avg: 453.79 / Max: 677.58Min: 101.92 / Avg: 460.64 / Max: 658.21

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-StreamSMT OnSMT Off110220330440550SE +/- 0.68, N = 3SE +/- 8.17, N = 15SE +/- 0.26, N = 3SE +/- 3.79, N = 3499.59468.78521.31426.74
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.09 / Avg: 235.18 / Max: 327.39Min: 20.67 / Avg: 228.97 / Max: 321.66Min: 42.89 / Avg: 453.79 / Max: 677.58Min: 101.92 / Avg: 460.64 / Max: 658.21
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-StreamSMT OnSMT Off90180270360450Min: 498.25 / Avg: 499.59 / Max: 500.51Min: 392.98 / Avg: 468.78 / Max: 496.26Min: 520.96 / Avg: 521.31 / Max: 521.81Min: 419.23 / Avg: 426.74 / Max: 431.39

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-StreamSMT OnSMT Off60120180240300SE +/- 0.15, N = 3SE +/- 2.61, N = 15SE +/- 0.12, N = 3SE +/- 2.51, N = 3126.93134.57242.96290.76
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-StreamSMT OnSMT Off50100150200250Min: 126.66 / Avg: 126.93 / Max: 127.19Min: 126.69 / Avg: 134.57 / Max: 159.51Min: 242.72 / Avg: 242.96 / Max: 243.1Min: 287.32 / Avg: 290.76 / Max: 295.64

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.98 / Avg: 231.96 / Max: 341.24Min: 20.52 / Avg: 236.78 / Max: 330.52Min: 42.65 / Avg: 496.12 / Max: 694.42Min: 192.96 / Avg: 504.41 / Max: 673.83

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-StreamSMT OnSMT Off60120180240300SE +/- 6.66, N = 15SE +/- 6.30, N = 15SE +/- 2.29, N = 15SE +/- 3.96, N = 15267.71259.28229.16212.68
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.98 / Avg: 231.96 / Max: 341.24Min: 20.52 / Avg: 236.78 / Max: 330.52Min: 42.65 / Avg: 496.12 / Max: 694.42Min: 192.96 / Avg: 504.41 / Max: 673.83
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-StreamSMT OnSMT Off50100150200250Min: 185.41 / Avg: 267.71 / Max: 285.21Min: 177.04 / Avg: 259.28 / Max: 274.27Min: 201.39 / Avg: 229.16 / Max: 236.59Min: 175.91 / Avg: 212.68 / Max: 231.11

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-StreamSMT OnSMT Off130260390520650SE +/- 7.96, N = 15SE +/- 8.20, N = 15SE +/- 6.10, N = 15SE +/- 11.89, N = 15240.87246.37556.94590.88
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-StreamSMT OnSMT Off100200300400500Min: 223.54 / Avg: 240.87 / Max: 343.89Min: 230.35 / Avg: 246.37 / Max: 357.1Min: 538.61 / Avg: 556.94 / Max: 632.45Min: 541.47 / Avg: 590.88 / Max: 710.54

PostgreSQL

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterPostgreSQL 15CPU Power Consumption MonitorSMT OnSMT Off70140210280350Min: 19.92 / Avg: 141 / Max: 209.56Min: 19.47 / Avg: 137.98 / Max: 261.97Min: 40.03 / Avg: 294.82 / Max: 404.71Min: 191.46 / Avg: 297.52 / Max: 385.38

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 15Scaling Factor: 1000 - Clients: 800 - Mode: Read Only - Average LatencySMT OnSMT Off0.22910.45820.68730.91641.1455SE +/- 0.040, N = 9SE +/- 0.020, N = 12SE +/- 0.025, N = 12SE +/- 0.006, N = 30.9520.9170.9741.0181. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterPostgreSQL 15CPU Power Consumption MonitorSMT OnSMT Off70140210280350Min: 19.92 / Avg: 141 / Max: 209.56Min: 19.47 / Avg: 137.98 / Max: 261.97Min: 40.03 / Avg: 294.82 / Max: 404.71Min: 191.46 / Avg: 297.52 / Max: 385.38
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 15Scaling Factor: 1000 - Clients: 800 - Mode: Read Only - Average LatencySMT OnSMT Off246810Min: 0.67 / Avg: 0.95 / Max: 1.08Min: 0.77 / Avg: 0.92 / Max: 0.98Min: 0.79 / Avg: 0.97 / Max: 1.06Min: 1.01 / Avg: 1.02 / Max: 1.031. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTPS, More Is BetterPostgreSQL 15Scaling Factor: 1000 - Clients: 800 - Mode: Read OnlySMT OnSMT Off200K400K600K800K1000KSE +/- 45813.37, N = 9SE +/- 21185.07, N = 12SE +/- 23693.25, N = 12SE +/- 4149.87, N = 38555698777228278787859681. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTPS, More Is BetterPostgreSQL 15Scaling Factor: 1000 - Clients: 800 - Mode: Read OnlySMT OnSMT Off150K300K450K600K750KMin: 743976.14 / Avg: 855569.48 / Max: 1199795.02Min: 814271.71 / Avg: 877721.86 / Max: 1045383.6Min: 752776.75 / Avg: 827877.81 / Max: 1012568.93Min: 778373.73 / Avg: 785967.6 / Max: 792665.231. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

Stockfish

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterStockfish 15CPU Power Consumption MonitorSMT OnSMT Off130260390520650Min: 21.02 / Avg: 326.55 / Max: 356.32Min: 20.41 / Avg: 309.55 / Max: 341.8Min: 42.5 / Avg: 649.7 / Max: 710.41Min: 41.98 / Avg: 598.99 / Max: 681.91

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgNodes Per Second Per Watt, More Is BetterStockfish 15Total TimeSMT OnSMT Off200K400K600K800K1000K1117842.33881021.76896399.09746293.69

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgNodes Per Second, More Is BetterStockfish 15Total TimeSMT OnSMT Off120M240M360M480M600MSE +/- 7021012.10, N = 12SE +/- 5762415.88, N = 15SE +/- 9265130.36, N = 15SE +/- 6859221.31, N = 123650343492727229405823869244470231431. (CXX) g++ options: -lgcov -m64 -lpthread -fno-exceptions -std=c++17 -fno-peel-loops -fno-tracer -pedantic -O3 -msse -msse3 -mpopcnt -mavx2 -mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl -msse4.1 -mssse3 -msse2 -mbmi2 -flto -flto=jobserver
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgNodes Per Second Per Watt, More Is BetterStockfish 15Total TimeSMT OnSMT Off200K400K600K800K1000K1117842.33881021.76896399.09746293.69
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterStockfish 15CPU Power Consumption MonitorSMT OnSMT Off130260390520650Min: 21.02 / Avg: 326.55 / Max: 356.32Min: 20.41 / Avg: 309.55 / Max: 341.8Min: 42.5 / Avg: 649.7 / Max: 710.41Min: 41.98 / Avg: 598.99 / Max: 681.91
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgNodes Per Second, More Is BetterStockfish 15Total TimeSMT OnSMT Off100M200M300M400M500MMin: 337283080 / Avg: 365034348.75 / Max: 407514704Min: 241060730 / Avg: 272722940.27 / Max: 308317378Min: 530039618 / Avg: 582386924.07 / Max: 632186487Min: 413873255 / Avg: 447023143.08 / Max: 4835260821. (CXX) g++ options: -lgcov -m64 -lpthread -fno-exceptions -std=c++17 -fno-peel-loops -fno-tracer -pedantic -O3 -msse -msse3 -mpopcnt -mavx2 -mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl -msse4.1 -mssse3 -msse2 -mbmi2 -flto -flto=jobserver

LuxCoreRender

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterLuxCoreRender 2.6CPU Power Consumption MonitorSMT OnSMT Off70140210280350Min: 20.9 / Avg: 142.31 / Max: 202.02Min: 20.14 / Avg: 136.45 / Max: 184.56Min: 41.26 / Avg: 281.59 / Max: 380.32Min: 40.47 / Avg: 279.04 / Max: 366.94

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgM samples/sec Per Watt, More Is BetterLuxCoreRender 2.6Scene: Rainbow Colors and Prism - Acceleration: CPUSMT OnSMT Off0.03310.06620.09930.13240.16550.1470.1050.0680.048

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: Rainbow Colors and Prism - Acceleration: CPUSMT OnSMT Off510152025SE +/- 0.03, N = 5SE +/- 0.04, N = 4SE +/- 0.03, N = 5SE +/- 0.35, N = 1520.8814.3719.0213.51
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgM samples/sec Per Watt, More Is BetterLuxCoreRender 2.6Scene: Rainbow Colors and Prism - Acceleration: CPUSMT OnSMT Off0.03310.06620.09930.13240.16550.1470.1050.0680.048
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterLuxCoreRender 2.6CPU Power Consumption MonitorSMT OnSMT Off70140210280350Min: 20.9 / Avg: 142.31 / Max: 202.02Min: 20.14 / Avg: 136.45 / Max: 184.56Min: 41.26 / Avg: 281.59 / Max: 380.32Min: 40.47 / Avg: 279.04 / Max: 366.94
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: Rainbow Colors and Prism - Acceleration: CPUSMT OnSMT Off510152025Min: 20.81 / Avg: 20.88 / Max: 20.97Min: 14.26 / Avg: 14.37 / Max: 14.45Min: 18.91 / Avg: 19.02 / Max: 19.08Min: 12.76 / Avg: 13.51 / Max: 16.62

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterLuxCoreRender 2.6CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.26 / Avg: 298.35 / Max: 337.28Min: 20.39 / Avg: 293.46 / Max: 329.9Min: 42.69 / Avg: 558.11 / Max: 673.06Min: 41.2 / Avg: 501.47 / Max: 651.41

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgM samples/sec Per Watt, More Is BetterLuxCoreRender 2.6Scene: Orange Juice - Acceleration: CPUSMT OnSMT Off0.01850.0370.05550.0740.09250.0820.0710.0620.050

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: Orange Juice - Acceleration: CPUSMT OnSMT Off816243240SE +/- 0.30, N = 15SE +/- 0.03, N = 3SE +/- 1.35, N = 15SE +/- 0.60, N = 1524.4720.9834.4525.11
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgM samples/sec Per Watt, More Is BetterLuxCoreRender 2.6Scene: Orange Juice - Acceleration: CPUSMT OnSMT Off0.01850.0370.05550.0740.09250.0820.0710.0620.050
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterLuxCoreRender 2.6CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 21.26 / Avg: 298.35 / Max: 337.28Min: 20.39 / Avg: 293.46 / Max: 329.9Min: 42.69 / Avg: 558.11 / Max: 673.06Min: 41.2 / Avg: 501.47 / Max: 651.41
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: Orange Juice - Acceleration: CPUSMT OnSMT Off714212835Min: 23.78 / Avg: 24.47 / Max: 26.76Min: 20.94 / Avg: 20.98 / Max: 21.04Min: 29.5 / Avg: 34.45 / Max: 41.29Min: 23.32 / Avg: 25.11 / Max: 29.82

srsRAN Project

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BettersrsRAN Project 23.5CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.98 / Avg: 261.42 / Max: 327.2Min: 20.63 / Avg: 260.57 / Max: 317.08Min: 43.09 / Avg: 560.78 / Max: 670.82Min: 41.48 / Avg: 517.72 / Max: 634.9

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgMbps Per Watt, More Is BettersrsRAN Project 23.5Test: PUSCH Processor Benchmark, Throughput TotalSMT OnSMT Off2040608010032.0978.4131.9070.64

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.5Test: PUSCH Processor Benchmark, Throughput TotalSMT OnSMT Off8K16K24K32K40KSE +/- 50.60, N = 3SE +/- 54.33, N = 3SE +/- 831.80, N = 15SE +/- 211.99, N = 38389.020430.917891.436573.81. (CXX) g++ options: -march=native -mfma -O3 -fno-trapping-math -fno-math-errno -lgtest
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgMbps Per Watt, More Is BettersrsRAN Project 23.5Test: PUSCH Processor Benchmark, Throughput TotalSMT OnSMT Off2040608010032.0978.4131.9070.64
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BettersrsRAN Project 23.5CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.98 / Avg: 261.42 / Max: 327.2Min: 20.63 / Avg: 260.57 / Max: 317.08Min: 43.09 / Avg: 560.78 / Max: 670.82Min: 41.48 / Avg: 517.72 / Max: 634.9
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.5Test: PUSCH Processor Benchmark, Throughput TotalSMT OnSMT Off6K12K18K24K30KMin: 8314.1 / Avg: 8389.03 / Max: 8485.4Min: 20350.3 / Avg: 20430.87 / Max: 20534.3Min: 16635.3 / Avg: 17891.35 / Max: 29112.9Min: 36236 / Avg: 36573.83 / Max: 36964.61. (CXX) g++ options: -march=native -mfma -O3 -fno-trapping-math -fno-math-errno -lgtest

Xmrig

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterXmrig 6.18.1CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.63 / Avg: 221.93 / Max: 329.48Min: 20.83 / Avg: 268.02 / Max: 330.69Min: 41.43 / Avg: 488.1 / Max: 682.33Min: 40.87 / Avg: 462.3 / Max: 653.34

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgH/s Per Watt, More Is BetterXmrig 6.18.1Variant: Monero - Hash Count: 1MSMT OnSMT Off4080120160200109.99191.10177.29185.91

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgH/s, More Is BetterXmrig 6.18.1Variant: Monero - Hash Count: 1MSMT OnSMT Off20K40K60K80K100KSE +/- 587.76, N = 15SE +/- 513.99, N = 3SE +/- 871.70, N = 4SE +/- 83.35, N = 424409.451218.986533.785946.01. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgH/s Per Watt, More Is BetterXmrig 6.18.1Variant: Monero - Hash Count: 1MSMT OnSMT Off4080120160200109.99191.10177.29185.91
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterXmrig 6.18.1CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.63 / Avg: 221.93 / Max: 329.48Min: 20.83 / Avg: 268.02 / Max: 330.69Min: 41.43 / Avg: 488.1 / Max: 682.33Min: 40.87 / Avg: 462.3 / Max: 653.34
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgH/s, More Is BetterXmrig 6.18.1Variant: Monero - Hash Count: 1MSMT OnSMT Off15K30K45K60K75KMin: 23549.4 / Avg: 24409.36 / Max: 32596.6Min: 50203.3 / Avg: 51218.9 / Max: 51864.5Min: 84196.3 / Avg: 86533.65 / Max: 88160.1Min: 85770.6 / Avg: 85945.98 / Max: 86169.81. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc

nekRS

nekRS is an open-source Navier Stokes solver based on the spectral element method. NekRS supports both CPU and GPU/accelerator support though this test profile is currently configured for CPU execution. NekRS is part of Nek5000 of the Mathematics and Computer Science MCS at Argonne National Laboratory. This nekRS benchmark is primarily relevant to large core count HPC servers and otherwise may be very time consuming on smaller systems. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgflops/rank, More Is BetternekRS 23.0Input: TurboPipe PeriodicSMT OnSMT Off600M1200M1800M2400M3000MSE +/- 62315945.32, N = 13SE +/- 87729075.59, N = 15253840692325860833331. (CXX) g++ options: -fopenmp -O2 -march=native -mtune=native -ftree-vectorize -rdynamic -lmpi_cxx -lmpi
OpenBenchmarking.orgflops/rank Per Watt, More Is BetternekRS 23.0Input: TurboPipe PeriodicSMT OnSMT Off2M4M6M8M10M8806106.198992432.84
OpenBenchmarking.orgWatts, Fewer Is BetternekRS 23.0CPU Power Consumption MonitorSMT OnSMT Off60120180240300Min: 21.11 / Avg: 288.26 / Max: 349.64Min: 20.65 / Avg: 287.58 / Max: 350.2
OpenBenchmarking.orgflops/rank, More Is BetternekRS 23.0Input: TurboPipe PeriodicSMT OnSMT Off400M800M1200M1600M2000MMin: 2417740000 / Avg: 2538406923.08 / Max: 3280590000Min: 2434380000 / Avg: 2586083333.33 / Max: 37924800001. (CXX) g++ options: -fopenmp -O2 -march=native -mtune=native -ftree-vectorize -rdynamic -lmpi_cxx -lmpi

Input: TurboPipe Periodic

EPYC 9754 2P: SMT On: The test quit with a non-zero exit status. E: mpirun noticed that process rank 0 with PID 0 on node phoronix-QuantaGrid-D54Q-2U exited on signal 11 (Segmentation fault).

EPYC 9754 2P: SMT Off: The test quit with a non-zero exit status. E: mpirun noticed that process rank 0 with PID 0 on node phoronix-QuantaGrid-D54Q-2U exited on signal 11 (Segmentation fault).

SPECFEM3D

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterSPECFEM3D 4.0CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.79 / Avg: 218.63 / Max: 337.15Min: 20.06 / Avg: 188.68 / Max: 325.98Min: 41.07 / Avg: 338.57 / Max: 663.27Min: 40.89 / Avg: 314 / Max: 658.8

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.0Model: Homogeneous HalfspaceSMT OnSMT Off3691215SE +/- 0.048294894, N = 4SE +/- 0.044108994, N = 15SE +/- 0.120152278, N = 15SE +/- 0.065109589, N = 129.4043485006.0927502994.7278301483.4518301691. (F9X) gfortran options: -O2 -fopenmp -std=f2003 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterSPECFEM3D 4.0CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.79 / Avg: 218.63 / Max: 337.15Min: 20.06 / Avg: 188.68 / Max: 325.98Min: 41.07 / Avg: 338.57 / Max: 663.27Min: 40.89 / Avg: 314 / Max: 658.8
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.0Model: Homogeneous HalfspaceSMT OnSMT Off3691215Min: 9.29 / Avg: 9.4 / Max: 9.51Min: 5.9 / Avg: 6.09 / Max: 6.54Min: 3.91 / Avg: 4.73 / Max: 5.26Min: 3.2 / Avg: 3.45 / Max: 3.991. (F9X) gfortran options: -O2 -fopenmp -std=f2003 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterSPECFEM3D 4.0CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.85 / Avg: 203.93 / Max: 337.24Min: 20.14 / Avg: 175.44 / Max: 323.55Min: 42.2 / Avg: 321.36 / Max: 664.42Min: 41.07 / Avg: 299.85 / Max: 655.45

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.0Model: Tomographic ModelSMT OnSMT Off246810SE +/- 0.080164897, N = 5SE +/- 0.048140615, N = 15SE +/- 0.086611165, N = 15SE +/- 0.014728964, N = 67.4806747064.9372655063.7827417982.7092054281. (F9X) gfortran options: -O2 -fopenmp -std=f2003 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterSPECFEM3D 4.0CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.85 / Avg: 203.93 / Max: 337.24Min: 20.14 / Avg: 175.44 / Max: 323.55Min: 42.2 / Avg: 321.36 / Max: 664.42Min: 41.07 / Avg: 299.85 / Max: 655.45
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.0Model: Tomographic ModelSMT OnSMT Off3691215Min: 7.24 / Avg: 7.48 / Max: 7.66Min: 4.75 / Avg: 4.94 / Max: 5.36Min: 3.33 / Avg: 3.78 / Max: 4.27Min: 2.67 / Avg: 2.71 / Max: 2.761. (F9X) gfortran options: -O2 -fopenmp -std=f2003 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

HeFFTe - Highly Efficient FFT for Exascale

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 20.59 / Avg: 233.52 / Max: 280.97Min: 20.21 / Avg: 229.75 / Max: 280.78Min: 43.55 / Avg: 408.39 / Max: 622.11Min: 41.26 / Avg: 404.72 / Max: 596.32

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOP/s Per Watt, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: r2c - Backend: FFTW - Precision: double - X Y Z: 512SMT OnSMT Off0.11720.23440.35160.46880.5860.2840.2960.5070.521

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: r2c - Backend: FFTW - Precision: double - X Y Z: 512SMT OnSMT Off50100150200250SE +/- 3.14, N = 15SE +/- 3.41, N = 15SE +/- 0.28, N = 5SE +/- 1.62, N = 566.3467.99207.20210.931. (CXX) g++ options: -O3
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOP/s Per Watt, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: r2c - Backend: FFTW - Precision: double - X Y Z: 512SMT OnSMT Off0.11720.23440.35160.46880.5860.2840.2960.5070.521
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 20.59 / Avg: 233.52 / Max: 280.97Min: 20.21 / Avg: 229.75 / Max: 280.78Min: 43.55 / Avg: 408.39 / Max: 622.11Min: 41.26 / Avg: 404.72 / Max: 596.32
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: r2c - Backend: FFTW - Precision: double - X Y Z: 512SMT OnSMT Off4080120160200Min: 61.38 / Avg: 66.34 / Max: 109.86Min: 62.09 / Avg: 67.99 / Max: 114.96Min: 206.48 / Avg: 207.2 / Max: 207.99Min: 205.32 / Avg: 210.93 / Max: 214.161. (CXX) g++ options: -O3

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 20.58 / Avg: 248.88 / Max: 282.9Min: 20.26 / Avg: 246.69 / Max: 279.29Min: 41.88 / Avg: 458.33 / Max: 608.28Min: 40.61 / Avg: 446.09 / Max: 590.49

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOP/s Per Watt, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: c2c - Backend: FFTW - Precision: double - X Y Z: 512SMT OnSMT Off0.05670.11340.17010.22680.28350.1400.1440.2390.252

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: c2c - Backend: FFTW - Precision: double - X Y Z: 512SMT OnSMT Off306090120150SE +/- 2.09, N = 15SE +/- 2.10, N = 15SE +/- 0.77, N = 3SE +/- 1.48, N = 334.8535.40109.65112.411. (CXX) g++ options: -O3
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOP/s Per Watt, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: c2c - Backend: FFTW - Precision: double - X Y Z: 512SMT OnSMT Off0.05670.11340.17010.22680.28350.1400.1440.2390.252
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 20.58 / Avg: 248.88 / Max: 282.9Min: 20.26 / Avg: 246.69 / Max: 279.29Min: 41.88 / Avg: 458.33 / Max: 608.28Min: 40.61 / Avg: 446.09 / Max: 590.49
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: c2c - Backend: FFTW - Precision: double - X Y Z: 512SMT OnSMT Off20406080100Min: 31.57 / Avg: 34.85 / Max: 63.91Min: 32.31 / Avg: 35.4 / Max: 64.5Min: 108.21 / Avg: 109.65 / Max: 110.83Min: 109.55 / Avg: 112.41 / Max: 114.491. (CXX) g++ options: -O3

libxsmm

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is Betterlibxsmm 2-1.17-3645CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.34 / Avg: 205.01 / Max: 330.64Min: 20.02 / Avg: 204.17 / Max: 323.63Min: 42.27 / Avg: 322.92 / Max: 664.29Min: 41.28 / Avg: 324.27 / Max: 648.28

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOPS/s Per Watt, More Is Betterlibxsmm 2-1.17-3645M N K: 128SMT OnSMT Off4812162013.2413.2115.4113.89

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOPS/s, More Is Betterlibxsmm 2-1.17-3645M N K: 128SMT OnSMT Off11002200330044005500SE +/- 0.99, N = 3SE +/- 19.26, N = 3SE +/- 62.14, N = 4SE +/- 114.55, N = 92713.42696.54976.74505.51. (CXX) g++ options: -dynamic -Bstatic -static-libgcc -lgomp -lm -lrt -ldl -lquadmath -lstdc++ -pthread -fPIC -std=c++14 -O2 -fopenmp-simd -funroll-loops -ftree-vectorize -fdata-sections -ffunction-sections -fvisibility=hidden -msse4.2
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOPS/s Per Watt, More Is Betterlibxsmm 2-1.17-3645M N K: 128SMT OnSMT Off4812162013.2413.2115.4113.89
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is Betterlibxsmm 2-1.17-3645CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.34 / Avg: 205.01 / Max: 330.64Min: 20.02 / Avg: 204.17 / Max: 323.63Min: 42.27 / Avg: 322.92 / Max: 664.29Min: 41.28 / Avg: 324.27 / Max: 648.28
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFLOPS/s, More Is Betterlibxsmm 2-1.17-3645M N K: 128SMT OnSMT Off9001800270036004500Min: 2712 / Avg: 2713.37 / Max: 2715.3Min: 2671.6 / Avg: 2696.5 / Max: 2734.4Min: 4863.4 / Avg: 4976.65 / Max: 5146.2Min: 3869.7 / Avg: 4505.46 / Max: 4924.11. (CXX) g++ options: -dynamic -Bstatic -static-libgcc -lgomp -lm -lrt -ldl -lquadmath -lstdc++ -pthread -fPIC -std=c++14 -O2 -fopenmp-simd -funroll-loops -ftree-vectorize -fdata-sections -ffunction-sections -fvisibility=hidden -msse4.2

miniBUDE

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterminiBUDE 20210901CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 20.2 / Avg: 155.92 / Max: 319.62Min: 19.77 / Avg: 154.01 / Max: 315.25Min: 40.84 / Avg: 269.38 / Max: 547.76Min: 40.2 / Avg: 223.4 / Max: 628.17

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgBillion Interactions/s Per Watt, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM1SMT OnSMT Off0.44210.88421.32631.76842.21051.5251.5240.9401.965

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgBillion Interactions/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM1SMT OnSMT Off100200300400500SE +/- 0.05, N = 9SE +/- 0.10, N = 9SE +/- 0.21, N = 9SE +/- 7.21, N = 15237.76234.68253.12439.071. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgBillion Interactions/s Per Watt, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM1SMT OnSMT Off0.44210.88421.32631.76842.21051.5251.5240.9401.965
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterminiBUDE 20210901CPU Power Consumption MonitorSMT OnSMT Off110220330440550Min: 20.2 / Avg: 155.92 / Max: 319.62Min: 19.77 / Avg: 154.01 / Max: 315.25Min: 40.84 / Avg: 269.38 / Max: 547.76Min: 40.2 / Avg: 223.4 / Max: 628.17
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgBillion Interactions/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM1SMT OnSMT Off80160240320400Min: 237.47 / Avg: 237.76 / Max: 237.98Min: 234.15 / Avg: 234.68 / Max: 235Min: 251.5 / Avg: 253.12 / Max: 253.57Min: 361.43 / Avg: 439.07 / Max: 459.261. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFInst/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM1SMT OnSMT Off2K4K6K8K10KSE +/- 1.30, N = 9SE +/- 2.53, N = 9SE +/- 5.35, N = 9SE +/- 180.32, N = 155944.065867.116328.0710976.681. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgGFInst/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM1SMT OnSMT Off2K4K6K8K10KMin: 5936.78 / Avg: 5944.06 / Max: 5949.56Min: 5853.62 / Avg: 5867.11 / Max: 5875.08Min: 6287.58 / Avg: 6328.07 / Max: 6339.15Min: 9035.77 / Avg: 10976.68 / Max: 11481.61. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

NAS Parallel Benchmarks

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNAS Parallel Benchmarks 3.4CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 19.9 / Avg: 127.5 / Max: 326.83Min: 19.52 / Avg: 119.85 / Max: 307.11Min: 41.07 / Avg: 228.71 / Max: 655.77Min: 39.68 / Avg: 227.42 / Max: 648.76

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s Per Watt, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: SP.BSMT OnSMT Off300600900120015001171.421347.261034.031082.90

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: SP.BSMT OnSMT Off50K100K150K200K250KSE +/- 1110.64, N = 9SE +/- 885.17, N = 9SE +/- 732.52, N = 9SE +/- 8884.57, N = 15149355.54161475.24236490.76246272.791. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s Per Watt, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: SP.BSMT OnSMT Off300600900120015001171.421347.261034.031082.90
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNAS Parallel Benchmarks 3.4CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 19.9 / Avg: 127.5 / Max: 326.83Min: 19.52 / Avg: 119.85 / Max: 307.11Min: 41.07 / Avg: 228.71 / Max: 655.77Min: 39.68 / Avg: 227.42 / Max: 648.76
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: SP.BSMT OnSMT Off40K80K120K160K200KMin: 143541.85 / Avg: 149355.54 / Max: 155024.55Min: 158221.02 / Avg: 161475.24 / Max: 164515.57Min: 233062.99 / Avg: 236490.76 / Max: 239111.27Min: 178205.97 / Avg: 246272.79 / Max: 281700.761. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNAS Parallel Benchmarks 3.4CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.22 / Avg: 238.61 / Max: 342.73Min: 19.6 / Avg: 229.32 / Max: 332.77Min: 41.42 / Avg: 399.82 / Max: 682.45Min: 40.26 / Avg: 375.87 / Max: 680.44

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s Per Watt, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: EP.DSMT OnSMT Off153045607555.5962.2559.2969.13

EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: EP.DSMT OnSMT Off6K12K18K24K30KSE +/- 214.86, N = 15SE +/- 54.02, N = 5SE +/- 413.32, N = 15SE +/- 940.02, N = 1213264.7914274.5323705.4025983.741. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s Per Watt, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: EP.DSMT OnSMT Off153045607555.5962.2559.2969.13
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgWatts, Fewer Is BetterNAS Parallel Benchmarks 3.4CPU Power Consumption MonitorSMT OnSMT Off120240360480600Min: 20.22 / Avg: 238.61 / Max: 342.73Min: 19.6 / Avg: 229.32 / Max: 332.77Min: 41.42 / Avg: 399.82 / Max: 682.45Min: 40.26 / Avg: 375.87 / Max: 680.44
EPYC 9754 1PEPYC 9754 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: EP.DSMT OnSMT Off5K10K15K20K25KMin: 11853.21 / Avg: 13264.79 / Max: 14371.32Min: 14148.51 / Avg: 14274.53 / Max: 14402.29Min: 21502.24 / Avg: 23705.4 / Max: 26559.87Min: 18784.44 / Avg: 25983.74 / Max: 28358.411. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2

184 Results Shown

OpenVINO
OpenSSL
toyBrot Fractal Generator
SPECFEM3D
7-Zip Compression
Neural Magic DeepSparse
John The Ripper:
  Blowfish
  bcrypt
Neural Magic DeepSparse:
  NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Stream
  CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Stream
  NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Stream
  NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Stream
Embree
Neural Magic DeepSparse
Graph500
OpenSSL
Embree
SPECFEM3D
Graph500
Blender
NAS Parallel Benchmarks
Liquid-DSP
OSPRay Studio:
  3 - 4K - 32 - Path Tracer
  3 - 4K - 16 - Path Tracer
  1 - 4K - 16 - Path Tracer
CP2K Molecular Dynamics
OSPRay Studio:
  1 - 4K - 1 - Path Tracer
  2 - 4K - 16 - Path Tracer
CloverLeaf
OSPRay Studio:
  3 - 4K - 1 - Path Tracer
  2 - 4K - 32 - Path Tracer
OpenSSL
OSPRay Studio
Graph500
OSPRay Studio
Blender
John The Ripper
Xmrig
ASTC Encoder
OpenVINO:
  Person Detection FP32 - CPU
  Machine Translation EN To DE FP16 - CPU
  Person Detection FP16 - CPU
SPECFEM3D
Blender:
  Barbershop - CPU-Only
  BMW27 - CPU-Only
Helsing
Graph500
OpenSSL
NAS Parallel Benchmarks
ASTC Encoder
OSPRay
OpenSSL
John The Ripper
OSPRay
Blender
OSPRay:
  particle_volume/ao/real_time
  gravity_spheres_volume/dim_512/scivis/real_time
OpenSSL
OpenVINO
OpenSSL:
  AES-128-GCM
  AES-256-GCM
OpenVINO:
  Machine Translation EN To DE FP16 - CPU
  Face Detection FP16-INT8 - CPU
  Face Detection FP16-INT8 - CPU
  Face Detection FP16 - CPU
Neural Magic DeepSparse
ASTC Encoder
OpenVINO:
  Weld Porosity Detection FP16-INT8 - CPU
  Weld Porosity Detection FP16 - CPU
NAMD
Liquid-DSP
OpenVINO
Timed Linux Kernel Compilation
OpenVINO
Primesieve
libxsmm
toyBrot Fractal Generator
OpenVINO
miniBUDE:
  OpenMP - BM2:
    GFInst/s
    Billion Interactions/s
NAS Parallel Benchmarks:
  IS.D
  BT.C
HeFFTe - Highly Efficient FFT for Exascale
NAS Parallel Benchmarks
LuxCoreRender
HeFFTe - Highly Efficient FFT for Exascale
Primesieve
OpenVINO
TensorFlow
Appleseed
NAS Parallel Benchmarks
7-Zip Compression
OpenVKL
TensorFlow
OpenVINO:
  Person Detection FP16 - CPU
  Person Detection FP32 - CPU
TensorFlow
OpenVINO
NAS Parallel Benchmarks
Timed Linux Kernel Compilation
LuxCoreRender
Neural Magic DeepSparse:
  NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Stream
  NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Stream
  CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Stream
  NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Stream
  NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Asynchronous Multi-Stream
Intel Open Image Denoise
Neural Magic DeepSparse:
  NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Stream
  CV Detection, YOLOv5s COCO - Asynchronous Multi-Stream
MariaDB
Appleseed
Intel Open Image Denoise
OpenVINO
Aircrack-ng
Intel Open Image Denoise
TensorFlow
MariaDB
Timed LLVM Compilation
TensorFlow
Timed Node.js Compilation
TensorFlow
miniFE
OpenVINO
Appleseed
Timed Gem5 Compilation
Timed LLVM Compilation
Timed Godot Game Engine Compilation
OpenVINO
nekRS
nekRS:
  CPU Power Consumption Monitor
  TurboPipe Periodic
  Phoronix Test Suite System Monitoring
  CPU Power Consumption Monitor
OpenVINO:
  Vehicle Detection FP16-INT8 - CPU:
    ms
    FPS
  Vehicle Detection FP16 - CPU:
    FPS
Neural Magic DeepSparse
Neural Magic DeepSparse:
  CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Stream:
    ms/batch
    items/sec
Neural Magic DeepSparse
Neural Magic DeepSparse:
  NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Stream:
    ms/batch
    items/sec
PostgreSQL
PostgreSQL:
  1000 - 800 - Read Only - Average Latency
  1000 - 800 - Read Only
Stockfish:
  CPU Power Consumption Monitor
  Total Time
Stockfish
LuxCoreRender:
  CPU Power Consumption Monitor
  Rainbow Colors and Prism - CPU
LuxCoreRender
LuxCoreRender:
  CPU Power Consumption Monitor
  Orange Juice - CPU
LuxCoreRender
srsRAN Project:
  CPU Power Consumption Monitor
  PUSCH Processor Benchmark, Throughput Total
srsRAN Project
Xmrig:
  CPU Power Consumption Monitor
  Monero - 1M
Xmrig
nekRS
SPECFEM3D
SPECFEM3D
SPECFEM3D
SPECFEM3D
HeFFTe - Highly Efficient FFT for Exascale:
  CPU Power Consumption Monitor
  r2c - FFTW - double - 512
HeFFTe - Highly Efficient FFT for Exascale
HeFFTe - Highly Efficient FFT for Exascale:
  CPU Power Consumption Monitor
  c2c - FFTW - double - 512
HeFFTe - Highly Efficient FFT for Exascale
libxsmm:
  CPU Power Consumption Monitor
  128
libxsmm
miniBUDE:
  CPU Power Consumption Monitor
  OpenMP - BM1
miniBUDE:
  OpenMP - BM1:
    Billion Interactions/s
    GFInst/s
NAS Parallel Benchmarks:
  CPU Power Consumption Monitor
  SP.B
NAS Parallel Benchmarks
NAS Parallel Benchmarks:
  CPU Power Consumption Monitor
  EP.D
NAS Parallel Benchmarks

EPYC 9754 1P: SMT On

Processor: AMD EPYC 9754 128-Core @ 2.25GHz (128 Cores / 256 Threads), Motherboard: AMD Titanite_4G (RTI1007B BIOS), Chipset: AMD Device 14a4, Memory: 768GB, Disk: 2 x 1920GB SAMSUNG MZWLJ1T9HBJR-00007, Graphics: ASPEED, Network: Broadcom NetXtreme BCM5720 PCIe

OS: Ubuntu 22.04, Kernel: 5.19.0-41-generic (x86_64), Desktop: GNOME Shell 42.5, Display Server: X Server 1.21.1.4, Vulkan: 1.3.224, Compiler: GCC 11.3.0, File-System: ext4, Screen Resolution: 1024x768

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-xKiWfi/gcc-11-11.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-xKiWfi/gcc-11-11.3.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v
Processor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xaa0010b
Python Notes: Python 3.10.6
Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 14 July 2023 05:22 by user phoronix.

EPYC 9754 1P: SMT Off

Processor: AMD EPYC 9754 128-Core @ 2.25GHz (128 Cores), Motherboard: AMD Titanite_4G (RTI1007B BIOS), Chipset: AMD Device 14a4, Memory: 768GB, Disk: 2 x 1920GB SAMSUNG MZWLJ1T9HBJR-00007, Graphics: ASPEED, Network: Broadcom NetXtreme BCM5720 PCIe

OS: Ubuntu 22.04, Kernel: 5.19.0-41-generic (x86_64), Desktop: GNOME Shell 42.5, Display Server: X Server 1.21.1.4, Vulkan: 1.3.224, Compiler: GCC 11.3.0, File-System: ext4, Screen Resolution: 1024x768

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-xKiWfi/gcc-11-11.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-xKiWfi/gcc-11-11.3.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v
Processor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xaa0010b
Python Notes: Python 3.10.6
Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: disabled RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 13 July 2023 10:55 by user phoronix.

EPYC 9754 2P: SMT On

Processor: 2 x AMD EPYC 9754 128-Core @ 2.25GHz (256 Cores / 512 Threads), Motherboard: AMD Titanite_4G (RTI1007B BIOS), Chipset: AMD Device 14a4, Memory: 1520GB, Disk: 2 x 1920GB SAMSUNG MZWLJ1T9HBJR-00007, Graphics: ASPEED, Network: Broadcom NetXtreme BCM5720 PCIe

OS: Ubuntu 22.04, Kernel: 5.19.0-41-generic (x86_64), Desktop: GNOME Shell 42.5, Display Server: X Server 1.21.1.4, Vulkan: 1.3.224, Compiler: GCC 11.3.0, File-System: ext4, Screen Resolution: 1024x768

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-xKiWfi/gcc-11-11.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-xKiWfi/gcc-11-11.3.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v
Processor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xaa0010b
Python Notes: Python 3.10.6
Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 11 July 2023 16:33 by user phoronix.

EPYC 9754 2P: SMT Off

Processor: 2 x AMD EPYC 9754 128-Core @ 2.25GHz (256 Cores), Motherboard: AMD Titanite_4G (RTI1007B BIOS), Chipset: AMD Device 14a4, Memory: 1520GB, Disk: 2 x 1920GB SAMSUNG MZWLJ1T9HBJR-00007, Graphics: ASPEED, Network: Broadcom NetXtreme BCM5720 PCIe

OS: Ubuntu 22.04, Kernel: 5.19.0-41-generic (x86_64), Desktop: GNOME Shell 42.5, Display Server: X Server 1.21.1.4, Vulkan: 1.3.224, Compiler: GCC 11.3.0, File-System: ext4, Screen Resolution: 1024x768

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-xKiWfi/gcc-11-11.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-xKiWfi/gcc-11-11.3.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v
Processor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xaa0010b
Python Notes: Python 3.10.6
Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: disabled RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 12 July 2023 08:31 by user phoronix.