AMD EPYC Turin 8c vs. 12c Memory Channel DDR5 Comparison

AMD EPYC 9655P memory benchmarks by Michael Larabel for a future article.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2411164-NE-AMDEPYCTU56
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
Show Result Confidence Charts

Limit displaying results to tests within:

C++ Boost Tests 2 Tests
Timed Code Compilation 3 Tests
C/C++ Compiler Tests 8 Tests
CPU Massive 18 Tests
Creator Workloads 14 Tests
Database Test Suite 4 Tests
Encoding 5 Tests
Fortran Tests 4 Tests
Game Development 4 Tests
HPC - High Performance Computing 17 Tests
Common Kernel Benchmarks 3 Tests
Machine Learning 5 Tests
Memory Test Suite 3 Tests
Molecular Dynamics 4 Tests
MPI Benchmarks 5 Tests
Multi-Core 22 Tests
NVIDIA GPU Compute 4 Tests
Intel oneAPI 3 Tests
OpenMPI Tests 10 Tests
Programmer / Developer System Benchmarks 4 Tests
Python Tests 6 Tests
Renderers 3 Tests
Scientific Computing 6 Tests
Software Defined Radio 2 Tests
Server 6 Tests
Server CPU Tests 11 Tests
Video Encoding 4 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
Prefer Vertical Bar Graphs

Additional Graphs

Show Perf Per RAM Channel Calculation Graphs Where Applicable

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Hide
Result
Result
Identifier
Performance Per
Dollar
Date
Run
  Test
  Duration
8c DDR5-6000
November 15
  14 Hours, 53 Minutes
12c DDR5-6000
November 15
  12 Hours, 2 Minutes
Invert Hiding All Results Option
  13 Hours, 27 Minutes
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):

8c DDR5-6000

Processor: AMD EPYC 9655P 96-Core @ 2.60GHz (96 Cores / 192 Threads), Motherboard: Supermicro Super Server H13SSL-N v1.01 (3.0 BIOS), Chipset: AMD 1Ah, Memory: 8 x 64GB DDR5-6000MT/s, Disk: 3201GB Micron_7450_MTFDKCB3T2TFS, Graphics: ASPEED, Network: 2 x Broadcom NetXtreme BCM5720 PCIe

OS: Ubuntu 24.10, Kernel: 6.12.0-rc7-phx (x86_64), Desktop: GNOME Shell 47.0, Display Server: X Server, Compiler: GCC 14.2.0, 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,d,fortran,objc,obj-c++,m2,rust --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-14-zdkDXv/gcc-14-14.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-14-zdkDXv/gcc-14-14.2.0/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: 0xb002116
Java Notes: OpenJDK Runtime Environment (build 21.0.5+11-Ubuntu-1ubuntu124.10)
Python Notes: Python 3.12.7
Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: 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 Enhanced / Automatic IBRS; IBPB: conditional; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected

12c DDR5-6000

Changed Memory to 12 x 64GB DDR5-6000MT/s Micron MTC40F2046S1RC64BDY QSFF.

AMD EPYC Turin 8c vs. 12c Memory Channel DDR5 ComparisonOpenBenchmarking.orgPhoronix Test SuiteAMD EPYC 9655P 96-Core @ 2.60GHz (96 Cores / 192 Threads)Supermicro Super Server H13SSL-N v1.01 (3.0 BIOS)AMD 1Ah8 x 64GB DDR5-6000MT/s12 x 64GB DDR5-6000MT/s Micron MTC40F2046S1RC64BDY QSFF3201GB Micron_7450_MTFDKCB3T2TFSASPEED2 x Broadcom NetXtreme BCM5720 PCIeUbuntu 24.106.12.0-rc7-phx (x86_64)GNOME Shell 47.0X ServerGCC 14.2.0ext41024x768ProcessorMotherboardChipsetMemoryDiskGraphicsNetworkOSKernelDesktopDisplay ServerCompilerFile-SystemScreen ResolutionAMD EPYC Turin 8c Vs. 12c Memory Channel DDR5 Comparison BenchmarksSystem Logs- Transparent Huge Pages: madvise- --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,d,fortran,objc,obj-c++,m2,rust --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-14-zdkDXv/gcc-14-14.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-14-zdkDXv/gcc-14-14.2.0/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 - Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xb002116- 8c DDR5-6000: OpenJDK Runtime Environment (build 21.0.5+11-Ubuntu-1ubuntu124.10)- Python 3.12.7- gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: 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 Enhanced / Automatic IBRS; IBPB: conditional; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected

8c DDR5-6000 vs. 12c DDR5-6000 ComparisonPhoronix Test SuiteBaseline+13.8%+13.8%+27.6%+27.6%+41.4%+41.4%55.1%53.9%53.9%53.5%52.5%51.7%51.4%50.6%47.5%43.5%42.7%42.3%39.8%37.2%32.3%27.2%26.8%23.4%23.3%22.9%22.8%20.9%20.7%18.5%18.4%18.2%18%17.6%17.4%17%16.1%14.2%14.2%13.4%11.5%11.3%10.4%8.7%8.5%8.2%8%7.9%7.9%7.6%7.3%7.3%7.2%6.8%6.7%6.6%6.6%6.4%6.3%6.2%5.9%5.7%4.9%4.4%4%3.5%3.3%3.3%3.2%2.8%2.7%2.7%2.4%2.3%2.2%2.6%9.2%3.3%6.7%4.4%CopyP.D.F - CPUScaleP.D.F - CPUAddTriad104 104 104 - 60144 144 144 - 60X.b.i.iSP.CChrysler Neon 1MMG.CFT.Cd.M.M.S - Execution Timei.i.1.C.P.DBT.CM.C.F.B.S - 8192 MiBIS.DCPU - 512 - ResNet-50M.C.F.B.S - 4096 MiBAdd - IntegerCopy - IntegerS.w.1.0.6.AScale - Integer1.R.H.D.S.RAverage - IntegerSP.B1.R.H.D.T.RTriad - Integer1.R.H.D.F.R.C.CM.T.E.T.D.F - CPUM.T.E.T.D.F - CPULU.C26Carbon Nanotube26CaffeNet 12-int8 - CPU - StandardN.S.P.L.F - CPU2626CG.CMobilenet QuantMemory Copy - 8192 MiBN.S.P.L.F - CPUMemory Copy - 4096 MiBCompression RatingEP.D7.1%MPI CPU - water_GMX50_bareZFNet-512 - CPU - Standard500MediumR.S.A.F.I - CPUR.S.A.F.I - CPUSmallWritesA.w.3.5.AConeGPT-2 - CPU - StandardLuxCore Benchmark - CPU4.3%FP32MobileNetV3Large5003.6%FastI.R.VT5 Encoder - CPU - StandardDeepLab V3DLSC - CPU3.2%d.S.M.S - Mesh Time2.9%fcn-resnet101-11 - CPU - Standardd.M.M.S - Mesh TimeBosphorus 4K - Super FastFP16MobileNetV3SmallTomographic Model2.3%v.IPreset 8 - Bosphorus 4KBosphorus 4K - Super Fast2%fcn-resnet101-11 - CPU - StandardCaffeNet 12-int8 - CPU - StandardT5 Encoder - CPU - StandardZFNet-512 - CPU - StandardGPT-2 - CPU - StandardStreamOpenVINOStreamOpenVINOStreamStreamHigh Performance Conjugate GradientHigh Performance Conjugate GradientAlgebraic Multi-Grid BenchmarkXcompact3d Incompact3dNAS Parallel BenchmarksOpenRadiossNAS Parallel BenchmarksNAS Parallel BenchmarksOpenFOAMXcompact3d Incompact3dNAS Parallel BenchmarksMBWNAS Parallel BenchmarksTensorFlowMBWRAMspeed SMPRAMspeed SMPNAMDRAMspeed SMPClickHouseRAMspeed SMPNAS Parallel BenchmarksClickHouseRAMspeed SMPClickHouseOpenVINOOpenVINONAS Parallel BenchmarksGraph500GPAWGraph500ONNX RuntimeOpenVINOGraph500Graph500NAS Parallel BenchmarksLiteRTMBWOpenVINOMBW7-Zip CompressionNAS Parallel BenchmarksGROMACSONNX RuntimeApache HTTP ServerWhisperfileOpenVINOOpenVINOWhisperfileApache CassandraNAMDEpochONNX RuntimeLuxCoreRenderXNNPACKnginxASTC EncoderLiteRTONNX RuntimeLiteRTLuxCoreRenderOpenFOAMONNX RuntimeOpenFOAMuvg266XNNPACKSPECFEM3DOpenVKLSVT-AV1KvazaarONNX RuntimeONNX RuntimeONNX RuntimeONNX RuntimeONNX Runtime8c DDR5-600012c DDR5-6000

AMD EPYC Turin 8c vs. 12c Memory Channel DDR5 Comparisonstream: Copyopenvino: Person Detection FP16 - CPUstream: Scaleopenvino: Person Detection FP16 - CPUstream: Addstream: Triadhpcg: 104 104 104 - 60hpcg: 144 144 144 - 60amg: incompact3d: X3D-benchmarking input.i3dnpb: SP.Copenradioss: Chrysler Neon 1Mnpb: MG.Cnpb: FT.Copenfoam: drivaerFastback, Medium Mesh Size - Execution Timeincompact3d: input.i3d 193 Cells Per Directionnpb: BT.Cmbw: Memory Copy, Fixed Block Size - 8192 MiBnpb: IS.Dtensorflow: CPU - 512 - ResNet-50mbw: Memory Copy, Fixed Block Size - 4096 MiBramspeed: Add - Integerramspeed: Copy - Integernamd: STMV with 1,066,628 Atomsramspeed: Scale - Integerclickhouse: 100M Rows Hits Dataset, Second Runramspeed: Average - Integernpb: SP.Bclickhouse: 100M Rows Hits Dataset, Third Runramspeed: Triad - Integerclickhouse: 100M Rows Hits Dataset, First Run / Cold Cacheopenvino: Machine Translation EN To DE FP16 - CPUopenvino: Machine Translation EN To DE FP16 - CPUnpb: LU.Cgraph500: 26gpaw: Carbon Nanotubegraph500: 26openvino: Noise Suppression Poconet-Like FP16 - CPUgraph500: 26graph500: 26npb: CG.Cmbw: Memory Copy - 8192 MiBopenvino: Noise Suppression Poconet-Like FP16 - CPUmbw: Memory Copy - 4096 MiBcompress-7zip: Compression Ratinggromacs: MPI CPU - water_GMX50_bareonnx: ZFNet-512 - CPU - Standardapache: 500whisperfile: Mediumopenvino: Road Segmentation ADAS FP16-INT8 - CPUopenvino: Road Segmentation ADAS FP16-INT8 - CPUwhisperfile: Smallcassandra: Writesnamd: ATPase with 327,506 Atomsepoch: Coneonnx: GPT-2 - CPU - Standardluxcorerender: LuxCore Benchmark - CPUxnnpack: FP32MobileNetV3Largenginx: 500astcenc: Fastlitert: Inception ResNet V2onnx: T5 Encoder - CPU - Standardlitert: DeepLab V3luxcorerender: DLSC - CPUopenfoam: drivaerFastback, Small Mesh Size - Mesh Timeonnx: fcn-resnet101-11 - CPU - Standardopenfoam: drivaerFastback, Medium Mesh Size - Mesh Timeuvg266: Bosphorus 4K - Super Fastxnnpack: FP16MobileNetV3Smallspecfem3d: Tomographic Modelopenvkl: vklBenchmarkCPU ISPCsvt-av1: Preset 8 - Bosphorus 4Kkvazaar: Bosphorus 4K - Super Fastliquid-dsp: 128 - 256 - 57onnx: bertsquad-12 - CPU - Standarduvg266: Bosphorus 4K - Very Fastbuild-linux-kernel: allmodconfigbuild-linux-kernel: defconfigvvenc: Bosphorus 4K - Fasterlitert: Quantized COCO SSD MobileNet v1pgbench: 1000 - 800 - Read Write - Average Latencyxnnpack: FP16MobileNetV3Largepgbench: 1000 - 800 - Read Writeluxcorerender: Rainbow Colors and Prism - CPUspecfem3d: Mount St. Helensnpb: EP.Connx: yolov4 - CPU - Standardxnnpack: QS8MobileNetV2john-the-ripper: MD5kvazaar: Bosphorus 4K - Very Fastxnnpack: FP16MobileNetV2memcached: 1:10uvg266: Bosphorus 4K - Ultra Fastspecfem3d: Layered Halfspaceonednn: Deconvolution Batch shapes_1d - CPUlitert: NASNet Mobilexnnpack: FP32MobileNetV3Smallwarpx: Uniform Plasmaopenfoam: drivaerFastback, Small Mesh Size - Execution Timesvt-av1: Preset 13 - Bosphorus 4Krawtherapee: Total Benchmark Timesrsran: PDSCH Processor Benchmark, Throughput Totallitert: Mobilenet Floatstargate: 192000 - 1024stargate: 192000 - 512onednn: Recurrent Neural Network Training - CPUmemcached: 1:100xnnpack: FP16MobileNetV1v-ray: CPUopenvino: Person Vehicle Bike Detection FP16 - CPUxnnpack: FP32MobileNetV1openradioss: INIVOL and Fluid Structure Interaction Drop Containerkvazaar: Bosphorus 4K - Ultra Fastopenvino: Person Vehicle Bike Detection FP16 - CPUonednn: Convolution Batch Shapes Auto - CPUsvt-av1: Preset 5 - Bosphorus 4Kliquid-dsp: 192 - 256 - 512liquid-dsp: 64 - 256 - 57specfem3d: Water-layered Halfspaceonednn: IP Shapes 1D - CPUminibude: OpenMP - BM2minibude: OpenMP - BM2svt-av1: Preset 3 - Bosphorus 4Kblender: Fishy Cat - CPU-Onlyvvenc: Bosphorus 4K - Fastonednn: IP Shapes 3D - CPUstargate: 96000 - 512build-godot: Time To Compilejohn-the-ripper: WPA PSKblender: BMW27 - CPU-Onlyopenvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPUuvg266: Bosphorus 4K - Mediumwarpx: Plasma Accelerationstargate: 96000 - 1024liquid-dsp: 64 - 256 - 512uvg266: Bosphorus 4K - Slowsvt-av1: Preset 3 - Beauty 4K 10-bitsvt-av1: Preset 8 - Beauty 4K 10-bitblender: Junkshop - CPU-Onlylitert: Inception V4memcached: 1:5blender: Barbershop - CPU-Onlyopenvino: Face Detection Retail FP16-INT8 - CPUluxcorerender: Danish Mood - CPUblender: Classroom - CPU-Onlyliquid-dsp: 128 - 256 - 512openvino: Handwritten English Recognition FP16-INT8 - CPUbuild2: Time To Compilem-queens: Time To Solvejohn-the-ripper: bcryptsvt-av1: Preset 13 - Beauty 4K 10-bitopenvino: Handwritten English Recognition FP16-INT8 - CPUpgbench: 1000 - 800 - Read Onlykvazaar: Bosphorus 4K - Mediumblender: Pabellon Barcelona - CPU-Onlyluxcorerender: Orange Juice - CPUsvt-av1: Preset 5 - Beauty 4K 10-bitopenvino: Weld Porosity Detection FP16-INT8 - CPUopenvino: Weld Porosity Detection FP16-INT8 - CPUonednn: Deconvolution Batch shapes_3d - CPUopenvino: Person Re-Identification Retail FP16 - CPUonednn: Recurrent Neural Network Inference - CPUastcenc: Thoroughcompress-7zip: Decompression Ratingastcenc: Mediumopenvino: Face Detection Retail FP16-INT8 - CPUkvazaar: Bosphorus 4K - Slowxnnpack: FP32MobileNetV2liquid-dsp: 192 - 256 - 57litert: SqueezeNetspecfem3d: Homogeneous Halfspaceopenvino: Face Detection FP16-INT8 - CPUjohn-the-ripper: HMAC-SHA512john-the-ripper: Blowfishastcenc: Very Thoroughopenvino: Vehicle Detection FP16-INT8 - CPUastcenc: Exhaustiveopenvino: Face Detection FP16-INT8 - CPUsrsran: PUSCH Processor Benchmark, Throughput Totalopenvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPUopenvino: Person Re-Identification Retail FP16 - CPUopenvino: Vehicle Detection FP16-INT8 - CPUpgbench: 1000 - 800 - Read Only - Average Latencyonnx: fcn-resnet101-11 - CPU - Standardonnx: CaffeNet 12-int8 - CPU - Standardonnx: CaffeNet 12-int8 - CPU - Standardonnx: bertsquad-12 - CPU - Standardonnx: T5 Encoder - CPU - Standardonnx: ZFNet-512 - CPU - Standardonnx: yolov4 - CPU - Standardonnx: GPT-2 - CPU - Standardlitert: Mobilenet Quantnpb: EP.D8c DDR5-600012c DDR5-6000221234.4106.96204677.8449.23233276.2229604.041.318641.01572085218000292.578786108558.21129.17105206.58120202.20254.295529.01416429284718.5318188.1115653.35196.4718332.40999666.50115580.473.55953117864.04652.53110436.65169436.81651.48104337.42637.6764.75740.13275011.4149779800028.5606767090007314.921365800000139445000060572.2023357.88912.4223448.63372704516.654153.637159864.99215.522892291.0920.8596.6879942231211.62457196.03187.12512.9214189523766.781072.781959977.8235.59611817.917.0122.2970729.18611109.1954365.52108106.6263693562762182.60893.40505320000021.647866.12196.24022.55522.6357110.127.1361325711217527.855.85535093311375.8312.060599682140833389.3390946752055.1266.8613.8612520996.84402100519.41053416.4537358022.648378430.86836.22826209.94441.274.3889743.896311429.68412031156.0345701981456329.60459679.9996.727.530.34137456.0882136766667321413333314.3553779040.5290527208.779288.35115.60318.559.5980.2655585.91859580.874101933313.09166363.9132.9225.762822726.517349141763333330.301.84613.51216.8043959.93711825.75121.153.9712.0635.5119563333333543.7953.5517.22923705517.11027.06314927745.3841.1824.157.6876.7014026.360.71536310448.92277.99682.1896626553568.336323443.8644.76923659554333337048.398.198005837330.1341332633323745811.63038240.987.1339145.116238.10.334.565.800.254109.0051.047505958.57646.20264.243726.5076182.94345.343596260.2114658.25343229.369.48314946.5689.50355858.4348347.362.544161.77223076286333203.937052154965.8490.77147097.41164903.45192.227217.08492692361153.0622449.8266970.89241.5522520.674120526.51139450.464.21717139548.85771.05130318.77199255.89765.07122052.34740.1556.71844.93311883.8055493600025.6607471380007934.401478180000150571000065380.5825126.49211.5725160.63277924217.787163.976170467.37202.159092437.9319.6191.0517644717612.28998186.85195.27612.3913643505484.861109.950158054.3243.32811450.916.4922.9446059.44076106.3117767.26105596.7779583002825186.68791.54496343333322.023867.24193.34622.22422.9607211.997.0371307411374528.235.77871461311230.8112.2149100952168100090.4689826670118.9467.6513.7000752296.7673699408.31041816.2745715222.408568435.38235.85425941.94399.484.4295703.932179425.78411923788.4346111997696380.58462979.4296.077.480.34359356.4382149200000319590000014.2755551360.5319847171.451286.85815.68418.469.6430.2667915.94397980.552102333313.14166983.3333.0425.851097726.539322141296666730.401.85213.55416.7544088.33701189.23120.833.9612.0335.4319521000003551.3553.4397.21423754217.14527.01315503645.3041.2524.197.6996.7114006.150.71637310462.56277.64382.2795627215568.826223427.5044.79924259586000007051.598.194862498330.0141347433323751511.62778240.407.1334145.106237.70.334.565.800.254106.21330.9595311041.9545.40674.108806.0971881.88165.119285800.6513685.97OpenBenchmarking.org

Stream

This is a benchmark of Stream, the popular system memory (RAM) benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterStream 2013-01-17Type: Copy12c DDR5-60008c DDR5-600070K140K210K280K350KSE +/- 570.76, N = 5SE +/- 357.86, N = 5343229.3221234.41. (CC) gcc options: -mcmodel=medium -O3 -march=native -fopenmp

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.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Person Detection FP16 - Device: CPU12c DDR5-60008c DDR5-600020406080100SE +/- 0.19, N = 3SE +/- 1.46, N = 1569.48106.96MIN: 34.01 / MAX: 134.55MIN: 37.41 / MAX: 380.851. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Stream

This is a benchmark of Stream, the popular system memory (RAM) benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterStream 2013-01-17Type: Scale12c DDR5-60008c DDR5-600070K140K210K280K350KSE +/- 478.12, N = 5SE +/- 630.12, N = 5314946.5204677.81. (CC) gcc options: -mcmodel=medium -O3 -march=native -fopenmp

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.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Person Detection FP16 - Device: CPU12c DDR5-60008c DDR5-6000150300450600750SE +/- 1.81, N = 3SE +/- 6.83, N = 15689.50449.231. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Stream

This is a benchmark of Stream, the popular system memory (RAM) benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterStream 2013-01-17Type: Add12c DDR5-60008c DDR5-600080K160K240K320K400KSE +/- 1639.97, N = 5SE +/- 1797.63, N = 5355858.4233276.21. (CC) gcc options: -mcmodel=medium -O3 -march=native -fopenmp

OpenBenchmarking.orgMB/s, More Is BetterStream 2013-01-17Type: Triad12c DDR5-60008c DDR5-600070K140K210K280K350KSE +/- 1895.03, N = 5SE +/- 936.98, N = 5348347.3229604.01. (CC) gcc options: -mcmodel=medium -O3 -march=native -fopenmp

High Performance Conjugate Gradient

HPCG is the High Performance Conjugate Gradient and is a new scientific benchmark from Sandia National Lans focused for super-computer testing with modern real-world workloads compared to HPCC. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHigh Performance Conjugate Gradient 3.1X Y Z: 104 104 104 - RT: 6012c DDR5-60008c DDR5-60001428425670SE +/- 0.03, N = 3SE +/- 0.02, N = 362.5441.321. (CXX) g++ options: -O3 -ffast-math -ftree-vectorize -lmpi_cxx -lmpi

OpenBenchmarking.orgGFLOP/s, More Is BetterHigh Performance Conjugate Gradient 3.1X Y Z: 144 144 144 - RT: 6012c DDR5-60008c DDR5-60001428425670SE +/- 0.01, N = 3SE +/- 0.03, N = 361.7741.021. (CXX) g++ options: -O3 -ffast-math -ftree-vectorize -lmpi_cxx -lmpi

Algebraic Multi-Grid Benchmark

AMG is a parallel algebraic multigrid solver for linear systems arising from problems on unstructured grids. The driver provided with AMG builds linear systems for various 3-dimensional problems. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFigure Of Merit, More Is BetterAlgebraic Multi-Grid Benchmark 1.212c DDR5-60008c DDR5-6000700M1400M2100M2800M3500MSE +/- 7144725.85, N = 3SE +/- 1073443.21, N = 330762863332085218000

Xcompact3d Incompact3d

Xcompact3d Incompact3d is a Fortran-MPI based, finite difference high-performance code for solving the incompressible Navier-Stokes equation and as many as you need scalar transport equations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterXcompact3d Incompact3d 2021-03-11Input: X3D-benchmarking input.i3d12c DDR5-60008c DDR5-600060120180240300SE +/- 2.76, N = 3SE +/- 3.79, N = 3203.94292.581. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

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.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: SP.C12c DDR5-60008c DDR5-600030K60K90K120K150KSE +/- 622.44, N = 3SE +/- 459.55, N = 3154965.84108558.211. (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 4.1.6

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: Chrysler Neon 1M12c DDR5-60008c DDR5-6000306090120150SE +/- 0.26, N = 3SE +/- 0.94, N = 1190.77129.17

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.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: MG.C12c DDR5-60008c DDR5-600030K60K90K120K150KSE +/- 2056.93, N = 3SE +/- 1082.06, N = 15147097.41105206.581. (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 4.1.6

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: FT.C12c DDR5-60008c DDR5-600040K80K120K160K200KSE +/- 1877.63, N = 4SE +/- 1124.88, N = 3164903.45120202.201. (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 4.1.6

OpenFOAM

OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Medium Mesh Size - Execution Time12c DDR5-60008c DDR5-600060120180240300192.23254.301. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfiniteVolume -lmeshTools -lparallel -llagrangian -lregionModels -lgenericPatchFields -lOpenFOAM -ldl -lm

Xcompact3d Incompact3d

Xcompact3d Incompact3d is a Fortran-MPI based, finite difference high-performance code for solving the incompressible Navier-Stokes equation and as many as you need scalar transport equations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterXcompact3d Incompact3d 2021-03-11Input: input.i3d 193 Cells Per Direction12c DDR5-60008c DDR5-60003691215SE +/- 0.06792326, N = 3SE +/- 0.05824182, N = 157.084926929.014164291. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

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.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: BT.C12c DDR5-60008c DDR5-600080K160K240K320K400KSE +/- 2078.86, N = 3SE +/- 377.10, N = 3361153.06284718.531. (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 4.1.6

MBW

This is a basic/simple memory (RAM) bandwidth benchmark for memory copy operations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterMBW 2018-09-08Test: Memory Copy, Fixed Block Size - Array Size: 8192 MiB12c DDR5-60008c DDR5-60005K10K15K20K25KSE +/- 64.32, N = 3SE +/- 60.40, N = 322449.8318188.111. (CC) gcc options: -O3 -march=native

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.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: IS.D12c DDR5-60008c DDR5-600015003000450060007500SE +/- 26.78, N = 3SE +/- 31.49, N = 36970.895653.351. (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 4.1.6

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.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 512 - Model: ResNet-5012c DDR5-60008c DDR5-600050100150200250SE +/- 0.11, N = 3SE +/- 0.27, N = 3241.55196.47

MBW

This is a basic/simple memory (RAM) bandwidth benchmark for memory copy operations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterMBW 2018-09-08Test: Memory Copy, Fixed Block Size - Array Size: 4096 MiB12c DDR5-60008c DDR5-60005K10K15K20K25KSE +/- 55.98, N = 3SE +/- 54.43, N = 322520.6718332.411. (CC) gcc options: -O3 -march=native

RAMspeed SMP

This benchmark tests the system memory (RAM) performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Type: Add - Benchmark: Integer12c DDR5-60008c DDR5-600030K60K90K120K150KSE +/- 235.46, N = 3SE +/- 1117.41, N = 3120526.5199666.501. (CC) gcc options: -O3 -march=native

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Type: Copy - Benchmark: Integer12c DDR5-60008c DDR5-600030K60K90K120K150KSE +/- 266.61, N = 3SE +/- 779.65, N = 3139450.46115580.471. (CC) gcc options: -O3 -march=native

NAMD

OpenBenchmarking.orgns/day, More Is BetterNAMD 3.0Input: STMV with 1,066,628 Atoms12c DDR5-60008c DDR5-60000.94891.89782.84673.79564.7445SE +/- 0.00356, N = 3SE +/- 0.00613, N = 34.217173.55953

RAMspeed SMP

This benchmark tests the system memory (RAM) performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Type: Scale - Benchmark: Integer12c DDR5-60008c DDR5-600030K60K90K120K150KSE +/- 283.77, N = 3SE +/- 291.38, N = 3139548.85117864.041. (CC) gcc options: -O3 -march=native

ClickHouse

ClickHouse is an open-source, high performance OLAP data management system. This test profile uses ClickHouse's standard benchmark recommendations per https://clickhouse.com/docs/en/operations/performance-test/ / https://github.com/ClickHouse/ClickBench/tree/main/clickhouse with the 100 million rows web analytics dataset. The reported value is the query processing time using the geometric mean of all separate queries performed as an aggregate. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Minute, Geo Mean, More Is BetterClickHouse 22.12.3.5100M Rows Hits Dataset, Second Run12c DDR5-60008c DDR5-6000170340510680850SE +/- 2.95, N = 3SE +/- 0.96, N = 3771.05652.53MIN: 69.77 / MAX: 8571.43MIN: 69.28 / MAX: 7500

RAMspeed SMP

This benchmark tests the system memory (RAM) performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Type: Average - Benchmark: Integer12c DDR5-60008c DDR5-600030K60K90K120K150KSE +/- 252.96, N = 3SE +/- 807.29, N = 3130318.77110436.651. (CC) gcc options: -O3 -march=native

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.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: SP.B12c DDR5-60008c DDR5-600040K80K120K160K200KSE +/- 2749.03, N = 3SE +/- 1524.43, N = 15199255.89169436.811. (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 4.1.6

ClickHouse

ClickHouse is an open-source, high performance OLAP data management system. This test profile uses ClickHouse's standard benchmark recommendations per https://clickhouse.com/docs/en/operations/performance-test/ / https://github.com/ClickHouse/ClickBench/tree/main/clickhouse with the 100 million rows web analytics dataset. The reported value is the query processing time using the geometric mean of all separate queries performed as an aggregate. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Minute, Geo Mean, More Is BetterClickHouse 22.12.3.5100M Rows Hits Dataset, Third Run12c DDR5-60008c DDR5-6000160320480640800SE +/- 4.74, N = 3SE +/- 1.81, N = 3765.07651.48MIN: 69.77 / MAX: 8571.43MIN: 70.92 / MAX: 6666.67

RAMspeed SMP

This benchmark tests the system memory (RAM) performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Type: Triad - Benchmark: Integer12c DDR5-60008c DDR5-600030K60K90K120K150KSE +/- 253.90, N = 3SE +/- 795.56, N = 3122052.34104337.421. (CC) gcc options: -O3 -march=native

ClickHouse

ClickHouse is an open-source, high performance OLAP data management system. This test profile uses ClickHouse's standard benchmark recommendations per https://clickhouse.com/docs/en/operations/performance-test/ / https://github.com/ClickHouse/ClickBench/tree/main/clickhouse with the 100 million rows web analytics dataset. The reported value is the query processing time using the geometric mean of all separate queries performed as an aggregate. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Minute, Geo Mean, More Is BetterClickHouse 22.12.3.5100M Rows Hits Dataset, First Run / Cold Cache12c DDR5-60008c DDR5-6000160320480640800SE +/- 3.26, N = 3SE +/- 1.98, N = 3740.15637.67MIN: 69.2 / MAX: 7500MIN: 67.95 / MAX: 7500

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.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Machine Translation EN To DE FP16 - Device: CPU12c DDR5-60008c DDR5-60001428425670SE +/- 0.09, N = 3SE +/- 0.40, N = 356.7164.75MIN: 34.97 / MAX: 96.61MIN: 28.81 / MAX: 99.111. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Machine Translation EN To DE FP16 - Device: CPU12c DDR5-60008c DDR5-60002004006008001000SE +/- 1.30, N = 3SE +/- 4.52, N = 3844.93740.131. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -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.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: LU.C12c DDR5-60008c DDR5-600070K140K210K280K350KSE +/- 2348.10, N = 3SE +/- 1080.76, N = 3311883.80275011.411. (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 4.1.6

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.

OpenBenchmarking.orgsssp median_TEPS, More Is BetterGraph500 3.0Scale: 2612c DDR5-60008c DDR5-6000120M240M360M480M600M5549360004977980001. (CC) gcc options: -fcommon -O3 -lpthread -lm -lmpi

GPAW

GPAW is a density-functional theory (DFT) Python code based on the projector-augmented wave (PAW) method and the atomic simulation environment (ASE). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterGPAW 23.6Input: Carbon Nanotube12c DDR5-60008c DDR5-6000714212835SE +/- 0.23, N = 3SE +/- 0.24, N = 325.6628.561. (CC) gcc options: -shared -fwrapv -O2 -lxc -lblas -lmpi

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.

OpenBenchmarking.orgsssp max_TEPS, More Is BetterGraph500 3.0Scale: 2612c DDR5-60008c DDR5-6000160M320M480M640M800M7471380006767090001. (CC) gcc options: -fcommon -O3 -lpthread -lm -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.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Noise Suppression Poconet-Like FP16 - Device: CPU12c DDR5-60008c DDR5-60002K4K6K8K10KSE +/- 20.16, N = 3SE +/- 10.57, N = 37934.407314.921. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -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.

OpenBenchmarking.orgbfs median_TEPS, More Is BetterGraph500 3.0Scale: 2612c DDR5-60008c DDR5-6000300M600M900M1200M1500M147818000013658000001. (CC) gcc options: -fcommon -O3 -lpthread -lm -lmpi

OpenBenchmarking.orgbfs max_TEPS, More Is BetterGraph500 3.0Scale: 2612c DDR5-60008c DDR5-6000300M600M900M1200M1500M150571000013944500001. (CC) gcc options: -fcommon -O3 -lpthread -lm -lmpi

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.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: CG.C12c DDR5-60008c DDR5-600014K28K42K56K70KSE +/- 928.22, N = 15SE +/- 631.87, N = 1565380.5860572.201. (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 4.1.6

MBW

This is a basic/simple memory (RAM) bandwidth benchmark for memory copy operations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterMBW 2018-09-08Test: Memory Copy - Array Size: 8192 MiB12c DDR5-60008c DDR5-60005K10K15K20K25KSE +/- 31.99, N = 3SE +/- 19.02, N = 325126.4923357.891. (CC) gcc options: -O3 -march=native

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.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Noise Suppression Poconet-Like FP16 - Device: CPU12c DDR5-60008c DDR5-60003691215SE +/- 0.03, N = 3SE +/- 0.02, N = 311.5712.42MIN: 6.95 / MAX: 30.79MIN: 6.11 / MAX: 28.551. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

MBW

This is a basic/simple memory (RAM) bandwidth benchmark for memory copy operations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterMBW 2018-09-08Test: Memory Copy - Array Size: 4096 MiB12c DDR5-60008c DDR5-60005K10K15K20K25KSE +/- 22.33, N = 3SE +/- 26.98, N = 325160.6323448.631. (CC) gcc options: -O3 -march=native

7-Zip Compression

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 24.05Test: Compression Rating12c DDR5-60008c DDR5-6000200K400K600K800K1000KSE +/- 879.82, N = 3SE +/- 1850.34, N = 37792427270451. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

GROMACS

The GROMACS (GROningen MAchine for Chemical Simulations) molecular dynamics package testing with the water_GMX50 data. This test profile allows selecting between CPU and GPU-based GROMACS builds. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNs Per Day, More Is BetterGROMACS 2024Implementation: MPI CPU - Input: water_GMX50_bare12c DDR5-60008c DDR5-600048121620SE +/- 0.02, N = 3SE +/- 0.03, N = 317.7916.651. (CXX) g++ options: -O3 -lm

ONNX Runtime

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: ZFNet-512 - Device: CPU - Executor: Standard12c DDR5-60008c DDR5-60004080120160200SE +/- 0.19, N = 3SE +/- 0.66, N = 3163.98153.641. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: ZFNet-512 - Device: CPU - Executor: Standard12c DDR5-60008c DDR5-6000246810SE +/- 0.00692, N = 3SE +/- 0.02791, N = 36.097186.507611. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

Apache HTTP Server

This is a test of the Apache HTTPD web server. This Apache HTTPD web server benchmark test profile makes use of the wrk program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.56Concurrent Requests: 50012c DDR5-60008c DDR5-600040K80K120K160K200KSE +/- 413.18, N = 3SE +/- 177.80, N = 3170467.37159864.991. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2

Whisperfile

OpenBenchmarking.orgSeconds, Fewer Is BetterWhisperfile 20Aug24Model Size: Medium12c DDR5-60008c DDR5-600050100150200250SE +/- 0.37, N = 3SE +/- 0.69, N = 3202.16215.52

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.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Road Segmentation ADAS FP16-INT8 - Device: CPU12c DDR5-60008c DDR5-60005001000150020002500SE +/- 1.15, N = 3SE +/- 2.70, N = 32437.932291.091. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Road Segmentation ADAS FP16-INT8 - Device: CPU12c DDR5-60008c DDR5-6000510152025SE +/- 0.01, N = 3SE +/- 0.02, N = 319.6120.85MIN: 10.37 / MAX: 41.25MIN: 10.51 / MAX: 40.511. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Whisperfile

OpenBenchmarking.orgSeconds, Fewer Is BetterWhisperfile 20Aug24Model Size: Small12c DDR5-60008c DDR5-600020406080100SE +/- 0.71, N = 3SE +/- 0.65, N = 391.0596.69

Apache Cassandra

OpenBenchmarking.orgOp/s, More Is BetterApache Cassandra 5.0Test: Writes12c DDR5-60008c DDR5-6000100K200K300K400K500KSE +/- 2454.69, N = 3SE +/- 1303.09, N = 3447176422312

NAMD

OpenBenchmarking.orgns/day, More Is BetterNAMD 3.0Input: ATPase with 327,506 Atoms12c DDR5-60008c DDR5-60003691215SE +/- 0.03, N = 3SE +/- 0.03, N = 312.2911.62

Epoch

OpenBenchmarking.orgSeconds, Fewer Is BetterEpoch 4.19.4Epoch3D Deck: Cone12c DDR5-60008c DDR5-60004080120160200SE +/- 0.55, N = 3SE +/- 0.52, N = 3186.85196.031. (F9X) gfortran options: -O3 -std=f2003 -Jobj -lsdf -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

ONNX Runtime

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: GPT-2 - Device: CPU - Executor: Standard12c DDR5-60008c DDR5-60004080120160200SE +/- 1.29, N = 3SE +/- 2.26, N = 3195.28187.131. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: GPT-2 - Device: CPU - Executor: Standard12c DDR5-60008c DDR5-60001.20232.40463.60694.80926.0115SE +/- 0.03369, N = 3SE +/- 0.06457, N = 35.119285.343591. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

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.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: LuxCore Benchmark - Acceleration: CPU8c DDR5-600012c DDR5-60003691215SE +/- 0.17, N = 3SE +/- 0.12, N = 1512.9212.39MIN: 6.3 / MAX: 15.04MIN: 5.62 / MAX: 14.88

XNNPACK

OpenBenchmarking.orgus, Fewer Is BetterXNNPACK b7b048Model: FP32MobileNetV3Large12c DDR5-60008c DDR5-60003K6K9K12K15KSE +/- 63.76, N = 3SE +/- 140.55, N = 313643141891. (CXX) g++ options: -O3 -lrt -lm

nginx

This is a benchmark of the lightweight Nginx HTTP(S) web-server. This Nginx web server benchmark test profile makes use of the wrk program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients/connections. HTTPS with a self-signed OpenSSL certificate is used by this test for local benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.23.2Connections: 5008c DDR5-600012c DDR5-6000110K220K330K440K550KSE +/- 3145.55, N = 3SE +/- 328.79, N = 3523766.78505484.861. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2

ASTC Encoder

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 5.0Preset: Fast12c DDR5-60008c DDR5-60002004006008001000SE +/- 14.18, N = 3SE +/- 7.37, N = 131109.951072.781. (CXX) g++ options: -O3 -flto -pthread

LiteRT

OpenBenchmarking.orgMicroseconds, Fewer Is BetterLiteRT 2024-10-15Model: Inception ResNet V212c DDR5-60008c DDR5-600013K26K39K52K65KSE +/- 598.16, N = 3SE +/- 811.71, N = 358054.359977.8

ONNX Runtime

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: T5 Encoder - Device: CPU - Executor: Standard12c DDR5-60008c DDR5-600050100150200250SE +/- 0.36, N = 3SE +/- 0.88, N = 3243.33235.601. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: T5 Encoder - Device: CPU - Executor: Standard12c DDR5-60008c DDR5-60000.95481.90962.86443.81924.774SE +/- 0.00593, N = 3SE +/- 0.01572, N = 34.108804.243721. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

LiteRT

OpenBenchmarking.orgMicroseconds, Fewer Is BetterLiteRT 2024-10-15Model: DeepLab V312c DDR5-60008c DDR5-60003K6K9K12K15KSE +/- 138.74, N = 3SE +/- 140.63, N = 1511450.911817.9

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.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: DLSC - Acceleration: CPU8c DDR5-600012c DDR5-600048121620SE +/- 0.21, N = 15SE +/- 0.19, N = 317.0116.49MIN: 15.67 / MAX: 20.15MIN: 15.79 / MAX: 20.03

OpenFOAM

OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Small Mesh Size - Mesh Time8c DDR5-600012c DDR5-600051015202522.3022.941. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfiniteVolume -lmeshTools -lparallel -llagrangian -lregionModels -lgenericPatchFields -lOpenFOAM -ldl -lm

ONNX Runtime

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: fcn-resnet101-11 - Device: CPU - Executor: Standard12c DDR5-60008c DDR5-60003691215SE +/- 0.13237, N = 15SE +/- 0.08964, N = 159.440769.186111. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: fcn-resnet101-11 - Device: CPU - Executor: Standard12c DDR5-60008c DDR5-600020406080100SE +/- 1.49, N = 15SE +/- 1.08, N = 15106.21109.011. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenFOAM

OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Medium Mesh Size - Mesh Time12c DDR5-60008c DDR5-600020406080100106.31109.201. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfiniteVolume -lmeshTools -lparallel -llagrangian -lregionModels -lgenericPatchFields -lOpenFOAM -ldl -lm

uvg266

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.8.0Video Input: Bosphorus 4K - Video Preset: Super Fast12c DDR5-60008c DDR5-60001530456075SE +/- 0.11, N = 3SE +/- 0.32, N = 367.2665.52

XNNPACK

OpenBenchmarking.orgus, Fewer Is BetterXNNPACK b7b048Model: FP16MobileNetV3Small12c DDR5-60008c DDR5-60002K4K6K8K10KSE +/- 181.63, N = 3SE +/- 162.45, N = 310559108101. (CXX) g++ options: -O3 -lrt -lm

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.

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.1.1Model: Tomographic Model8c DDR5-600012c DDR5-6000246810SE +/- 0.080343793, N = 3SE +/- 0.086144824, N = 36.6263693566.7779583001. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -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

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.

OpenBenchmarking.orgItems / Sec, More Is BetterOpenVKL 2.0.0Benchmark: vklBenchmarkCPU ISPC12c DDR5-60008c DDR5-60006001200180024003000SE +/- 1.76, N = 3SE +/- 1.53, N = 328252762MIN: 217 / MAX: 36373MIN: 216 / MAX: 33882

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.3Encoder Mode: Preset 8 - Input: Bosphorus 4K12c DDR5-60008c DDR5-60004080120160200SE +/- 1.50, N = 3SE +/- 0.19, N = 3186.69182.611. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Super Fast8c DDR5-600012c DDR5-600020406080100SE +/- 0.53, N = 3SE +/- 0.52, N = 393.4091.541. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

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.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 128 - Buffer Length: 256 - Filter Length: 578c DDR5-600012c DDR5-60001100M2200M3300M4400M5500MSE +/- 8868107.65, N = 3SE +/- 8326730.72, N = 3505320000049634333331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

ONNX Runtime

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: bertsquad-12 - Device: CPU - Executor: Standard12c DDR5-60008c DDR5-6000510152025SE +/- 0.12, N = 3SE +/- 0.23, N = 322.0221.651. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: bertsquad-12 - Device: CPU - Executor: Standard12c DDR5-60008c DDR5-60001020304050SE +/- 0.24, N = 3SE +/- 0.48, N = 345.4146.201. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

uvg266

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.8.0Video Input: Bosphorus 4K - Video Preset: Very Fast12c DDR5-60008c DDR5-60001530456075SE +/- 0.21, N = 3SE +/- 0.33, N = 367.2466.12

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.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.8Build: allmodconfig12c DDR5-60008c DDR5-60004080120160200SE +/- 0.43, N = 3SE +/- 0.13, N = 3193.35196.24

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.8Build: defconfig12c DDR5-60008c DDR5-6000510152025SE +/- 0.25, N = 4SE +/- 0.25, N = 422.2222.56

VVenC

VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVVenC 1.11Video Input: Bosphorus 4K - Video Preset: Faster12c DDR5-60008c DDR5-6000612182430SE +/- 0.06, N = 3SE +/- 0.02, N = 322.9622.641. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects

LiteRT

OpenBenchmarking.orgMicroseconds, Fewer Is BetterLiteRT 2024-10-15Model: Quantized COCO SSD MobileNet v18c DDR5-600012c DDR5-600015003000450060007500SE +/- 68.84, N = 15SE +/- 28.16, N = 37110.127211.99

PostgreSQL

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 17Scaling Factor: 1000 - Clients: 800 - Mode: Read Write - Average Latency12c DDR5-60008c DDR5-6000246810SE +/- 0.079, N = 5SE +/- 0.053, N = 117.0377.1361. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm

XNNPACK

OpenBenchmarking.orgus, Fewer Is BetterXNNPACK b7b048Model: FP16MobileNetV3Large12c DDR5-60008c DDR5-60003K6K9K12K15KSE +/- 58.09, N = 3SE +/- 75.41, N = 313074132571. (CXX) g++ options: -O3 -lrt -lm

PostgreSQL

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 17Scaling Factor: 1000 - Clients: 800 - Mode: Read Write12c DDR5-60008c DDR5-600020K40K60K80K100KSE +/- 1257.14, N = 5SE +/- 836.14, N = 111137451121751. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm

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.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: Rainbow Colors and Prism - Acceleration: CPU12c DDR5-60008c DDR5-6000714212835SE +/- 0.19, N = 3SE +/- 0.02, N = 328.2327.85MIN: 25.05 / MAX: 28.69MIN: 24.8 / MAX: 28.48

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.

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.1.1Model: Mount St. Helens12c DDR5-60008c DDR5-60001.31752.6353.95255.276.5875SE +/- 0.055041694, N = 6SE +/- 0.034598254, N = 35.7787146135.8553509331. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -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

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.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: EP.C8c DDR5-600012c DDR5-60002K4K6K8K10KSE +/- 91.88, N = 3SE +/- 85.90, N = 311375.8311230.811. (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 4.1.6

ONNX Runtime

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: yolov4 - Device: CPU - Executor: Standard12c DDR5-60008c DDR5-60003691215SE +/- 0.13, N = 3SE +/- 0.16, N = 312.2112.061. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: yolov4 - Device: CPU - Executor: Standard12c DDR5-60008c DDR5-600020406080100SE +/- 0.85, N = 3SE +/- 1.14, N = 381.8882.941. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

XNNPACK

OpenBenchmarking.orgus, Fewer Is BetterXNNPACK b7b048Model: QS8MobileNetV28c DDR5-600012c DDR5-60002K4K6K8K10KSE +/- 103.51, N = 3SE +/- 60.71, N = 39968100951. (CXX) g++ options: -O3 -lrt -lm

John The Ripper

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

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: MD512c DDR5-60008c DDR5-60005M10M15M20M25MSE +/- 20808.65, N = 3SE +/- 32518.37, N = 321681000214083331. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Very Fast12c DDR5-60008c DDR5-600020406080100SE +/- 0.61, N = 3SE +/- 0.20, N = 390.4689.331. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

XNNPACK

OpenBenchmarking.orgus, Fewer Is BetterXNNPACK b7b048Model: FP16MobileNetV212c DDR5-60008c DDR5-60002K4K6K8K10KSE +/- 96.09, N = 3SE +/- 49.08, N = 3898290941. (CXX) g++ options: -O3 -lrt -lm

Memcached

Memcached is a high performance, distributed memory object caching system. This Memcached test profiles makes use of memtier_benchmark for excuting this CPU/memory-focused server benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/sec, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:108c DDR5-600012c DDR5-60001.4M2.8M4.2M5.6M7MSE +/- 5728.08, N = 3SE +/- 23004.60, N = 36752055.126670118.941. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

uvg266

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.8.0Video Input: Bosphorus 4K - Video Preset: Ultra Fast12c DDR5-60008c DDR5-60001530456075SE +/- 0.18, N = 3SE +/- 0.79, N = 367.6566.86

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.

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.1.1Model: Layered Halfspace12c DDR5-60008c DDR5-600048121620SE +/- 0.07, N = 3SE +/- 0.11, N = 313.7013.861. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -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

oneDNN

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.6Harness: Deconvolution Batch shapes_1d - Engine: CPU12c DDR5-60008c DDR5-6000246810SE +/- 0.02120, N = 3SE +/- 0.01476, N = 36.767366.84402MIN: 4.07MIN: 5.141. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -fcf-protection=full -pie -ldl

LiteRT

OpenBenchmarking.orgMicroseconds, Fewer Is BetterLiteRT 2024-10-15Model: NASNet Mobile12c DDR5-60008c DDR5-600020K40K60K80K100KSE +/- 1271.15, N = 3SE +/- 1096.54, N = 1599408.3100519.4

XNNPACK

OpenBenchmarking.orgus, Fewer Is BetterXNNPACK b7b048Model: FP32MobileNetV3Small12c DDR5-60008c DDR5-60002K4K6K8K10KSE +/- 54.11, N = 3SE +/- 8.95, N = 310418105341. (CXX) g++ options: -O3 -lrt -lm

WarpX

OpenBenchmarking.orgSeconds, Fewer Is BetterWarpX 24.10Input: Uniform Plasma12c DDR5-60008c DDR5-600048121620SE +/- 0.23, N = 15SE +/- 0.24, N = 1216.2716.451. (CXX) g++ options: -O3 -lm

OpenFOAM

OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Small Mesh Size - Execution Time12c DDR5-60008c DDR5-600051015202522.4122.651. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfiniteVolume -lmeshTools -lparallel -llagrangian -lregionModels -lgenericPatchFields -lOpenFOAM -ldl -lm

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.3Encoder Mode: Preset 13 - Input: Bosphorus 4K12c DDR5-60008c DDR5-600090180270360450SE +/- 1.17, N = 3SE +/- 0.57, N = 3435.38430.871. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

RawTherapee

RawTherapee is a cross-platform, open-source multi-threaded RAW image processing program. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterRawTherapeeTotal Benchmark Time12c DDR5-60008c DDR5-6000816243240SE +/- 0.05, N = 3SE +/- 0.05, N = 335.8536.231. RawTherapee, version 5.10, command line.

srsRAN Project

srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.10.1-20240325Test: PDSCH Processor Benchmark, Throughput Total8c DDR5-600012c DDR5-60006K12K18K24K30KSE +/- 185.11, N = 12SE +/- 311.92, N = 326209.925941.91. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -mavx512f -mavx512cd -mavx512bw -mavx512dq -fno-trapping-math -fno-math-errno -ldl

LiteRT

OpenBenchmarking.orgMicroseconds, Fewer Is BetterLiteRT 2024-10-15Model: Mobilenet Float12c DDR5-60008c DDR5-600010002000300040005000SE +/- 16.88, N = 3SE +/- 18.31, N = 34399.484441.27

Stargate Digital Audio Workstation

Stargate is an open-source, cross-platform digital audio workstation (DAW) software package with "a unique and carefully curated experience" with scalability from old systems up through modern multi-core systems. Stargate is GPLv3 licensed and makes use of Qt5 (PyQt5) for its user-interface. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRender Ratio, More Is BetterStargate Digital Audio Workstation 22.11.5Sample Rate: 192000 - Buffer Size: 102412c DDR5-60008c DDR5-60000.99671.99342.99013.98684.9835SE +/- 0.018753, N = 3SE +/- 0.014254, N = 34.4295704.3889741. (CXX) g++ options: -lpthread -lsndfile -lm -O3 -march=native -ffast-math -funroll-loops -fstrength-reduce -fstrict-aliasing -finline-functions

OpenBenchmarking.orgRender Ratio, More Is BetterStargate Digital Audio Workstation 22.11.5Sample Rate: 192000 - Buffer Size: 51212c DDR5-60008c DDR5-60000.88471.76942.65413.53884.4235SE +/- 0.003414, N = 3SE +/- 0.040462, N = 33.9321793.8963111. (CXX) g++ options: -lpthread -lsndfile -lm -O3 -march=native -ffast-math -funroll-loops -fstrength-reduce -fstrict-aliasing -finline-functions

oneDNN

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.6Harness: Recurrent Neural Network Training - Engine: CPU12c DDR5-60008c DDR5-600090180270360450SE +/- 0.20, N = 3SE +/- 0.28, N = 3425.78429.68MIN: 420.18MIN: 423.831. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -fcf-protection=full -pie -ldl

Memcached

Memcached is a high performance, distributed memory object caching system. This Memcached test profiles makes use of memtier_benchmark for excuting this CPU/memory-focused server benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/sec, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:1008c DDR5-600012c DDR5-60003M6M9M12M15MSE +/- 48609.95, N = 3SE +/- 20796.91, N = 312031156.0311923788.431. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

XNNPACK

OpenBenchmarking.orgus, Fewer Is BetterXNNPACK b7b048Model: FP16MobileNetV18c DDR5-600012c DDR5-600010002000300040005000SE +/- 11.46, N = 3SE +/- 43.59, N = 3457046111. (CXX) g++ options: -O3 -lrt -lm

Chaos Group V-RAY

This is a test of Chaos Group's V-RAY benchmark. V-RAY is a commercial renderer that can integrate with various creator software products like SketchUp and 3ds Max. The V-RAY benchmark is standalone and supports CPU and NVIDIA CUDA/RTX based rendering. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgvsamples, More Is BetterChaos Group V-RAY 6.0Mode: CPU12c DDR5-60008c DDR5-600040K80K120K160K200KSE +/- 252.42, N = 3SE +/- 357.23, N = 3199769198145

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.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Person Vehicle Bike Detection FP16 - Device: CPU12c DDR5-60008c DDR5-600014002800420056007000SE +/- 12.38, N = 3SE +/- 17.72, N = 36380.586329.601. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

XNNPACK

OpenBenchmarking.orgus, Fewer Is BetterXNNPACK b7b048Model: FP32MobileNetV18c DDR5-600012c DDR5-600010002000300040005000SE +/- 41.29, N = 3SE +/- 57.47, N = 3459646291. (CXX) g++ options: -O3 -lrt -lm

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: INIVOL and Fluid Structure Interaction Drop Container12c DDR5-60008c DDR5-600020406080100SE +/- 0.06, N = 3SE +/- 0.30, N = 379.4279.99

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Ultra Fast8c DDR5-600012c DDR5-600020406080100SE +/- 0.36, N = 3SE +/- 0.29, N = 396.7296.071. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

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.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Person Vehicle Bike Detection FP16 - Device: CPU12c DDR5-60008c DDR5-6000246810SE +/- 0.01, N = 3SE +/- 0.02, N = 37.487.53MIN: 4.42 / MAX: 23.35MIN: 4.35 / MAX: 22.921. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

oneDNN

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.6Harness: Convolution Batch Shapes Auto - Engine: CPU8c DDR5-600012c DDR5-60000.07730.15460.23190.30920.3865SE +/- 0.000677, N = 3SE +/- 0.000927, N = 30.3413740.343593MIN: 0.33MIN: 0.321. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -fcf-protection=full -pie -ldl

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.3Encoder Mode: Preset 5 - Input: Bosphorus 4K12c DDR5-60008c DDR5-60001326395265SE +/- 0.28, N = 3SE +/- 0.10, N = 356.4456.091. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

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.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 192 - Buffer Length: 256 - Filter Length: 51212c DDR5-60008c DDR5-6000500M1000M1500M2000M2500MSE +/- 1582192.57, N = 3SE +/- 821245.67, N = 3214920000021367666671. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 64 - Buffer Length: 256 - Filter Length: 578c DDR5-600012c DDR5-6000700M1400M2100M2800M3500MSE +/- 3636084.59, N = 3SE +/- 1814754.35, N = 3321413333331959000001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

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.

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.1.1Model: Water-layered Halfspace12c DDR5-60008c DDR5-600048121620SE +/- 0.08, N = 3SE +/- 0.07, N = 314.2814.361. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -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

oneDNN

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.6Harness: IP Shapes 1D - Engine: CPU8c DDR5-600012c DDR5-60000.11970.23940.35910.47880.5985SE +/- 0.004111, N = 3SE +/- 0.003116, N = 30.5290520.531984MIN: 0.47MIN: 0.481. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -fcf-protection=full -pie -ldl

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.

OpenBenchmarking.orgGFInst/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM28c DDR5-600012c DDR5-600015003000450060007500SE +/- 12.59, N = 3SE +/- 56.00, N = 37208.787171.451. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

OpenBenchmarking.orgBillion Interactions/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM28c DDR5-600012c DDR5-600060120180240300SE +/- 0.50, N = 3SE +/- 2.24, N = 3288.35286.861. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.3Encoder Mode: Preset 3 - Input: Bosphorus 4K12c DDR5-60008c DDR5-600048121620SE +/- 0.03, N = 3SE +/- 0.00, N = 315.6815.601. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.2Blend File: Fishy Cat - Compute: CPU-Only12c DDR5-60008c DDR5-6000510152025SE +/- 0.05, N = 3SE +/- 0.03, N = 318.4618.55

VVenC

VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVVenC 1.11Video Input: Bosphorus 4K - Video Preset: Fast12c DDR5-60008c DDR5-60003691215SE +/- 0.082, N = 3SE +/- 0.064, N = 39.6439.5981. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects

oneDNN

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.6Harness: IP Shapes 3D - Engine: CPU8c DDR5-600012c DDR5-60000.060.120.180.240.3SE +/- 0.000649, N = 3SE +/- 0.000721, N = 30.2655580.266791MIN: 0.25MIN: 0.251. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -fcf-protection=full -pie -ldl

Stargate Digital Audio Workstation

Stargate is an open-source, cross-platform digital audio workstation (DAW) software package with "a unique and carefully curated experience" with scalability from old systems up through modern multi-core systems. Stargate is GPLv3 licensed and makes use of Qt5 (PyQt5) for its user-interface. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRender Ratio, More Is BetterStargate Digital Audio Workstation 22.11.5Sample Rate: 96000 - Buffer Size: 51212c DDR5-60008c DDR5-60001.33742.67484.01225.34966.687SE +/- 0.018441, N = 3SE +/- 0.033577, N = 35.9439795.9185951. (CXX) g++ options: -lpthread -lsndfile -lm -O3 -march=native -ffast-math -funroll-loops -fstrength-reduce -fstrict-aliasing -finline-functions

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.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Godot Game Engine Compilation 4.0Time To Compile12c DDR5-60008c DDR5-600020406080100SE +/- 0.19, N = 3SE +/- 0.04, N = 380.5580.87

John The Ripper

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

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: WPA PSK12c DDR5-60008c DDR5-6000200K400K600K800K1000KSE +/- 2185.81, N = 3SE +/- 1666.67, N = 3102333310193331. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.2Blend File: BMW27 - Compute: CPU-Only8c DDR5-600012c DDR5-60003691215SE +/- 0.05, N = 3SE +/- 0.05, N = 313.0913.14

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.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPU12c DDR5-60008c DDR5-600040K80K120K160K200KSE +/- 251.97, N = 3SE +/- 234.32, N = 3166983.33166363.911. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

uvg266

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.8.0Video Input: Bosphorus 4K - Video Preset: Medium12c DDR5-60008c DDR5-6000816243240SE +/- 0.06, N = 3SE +/- 0.02, N = 333.0432.92

WarpX

OpenBenchmarking.orgSeconds, Fewer Is BetterWarpX 24.10Input: Plasma Acceleration8c DDR5-600012c DDR5-6000612182430SE +/- 0.17, N = 3SE +/- 0.22, N = 325.7625.851. (CXX) g++ options: -O3 -lm

Stargate Digital Audio Workstation

Stargate is an open-source, cross-platform digital audio workstation (DAW) software package with "a unique and carefully curated experience" with scalability from old systems up through modern multi-core systems. Stargate is GPLv3 licensed and makes use of Qt5 (PyQt5) for its user-interface. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRender Ratio, More Is BetterStargate Digital Audio Workstation 22.11.5Sample Rate: 96000 - Buffer Size: 102412c DDR5-60008c DDR5-6000246810SE +/- 0.024623, N = 3SE +/- 0.002866, N = 36.5393226.5173491. (CXX) g++ options: -lpthread -lsndfile -lm -O3 -march=native -ffast-math -funroll-loops -fstrength-reduce -fstrict-aliasing -finline-functions

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.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 64 - Buffer Length: 256 - Filter Length: 5128c DDR5-600012c DDR5-6000300M600M900M1200M1500MSE +/- 2356079.61, N = 3SE +/- 1419311.26, N = 3141763333314129666671. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

uvg266

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.8.0Video Input: Bosphorus 4K - Video Preset: Slow12c DDR5-60008c DDR5-6000714212835SE +/- 0.01, N = 3SE +/- 0.04, N = 330.4030.30

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.3Encoder Mode: Preset 3 - Input: Beauty 4K 10-bit12c DDR5-60008c DDR5-60000.41670.83341.25011.66682.0835SE +/- 0.007, N = 3SE +/- 0.002, N = 31.8521.8461. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.3Encoder Mode: Preset 8 - Input: Beauty 4K 10-bit12c DDR5-60008c DDR5-60003691215SE +/- 0.00, N = 3SE +/- 0.02, N = 313.5513.511. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.2Blend File: Junkshop - Compute: CPU-Only12c DDR5-60008c DDR5-600048121620SE +/- 0.01, N = 3SE +/- 0.02, N = 316.7516.80

LiteRT

OpenBenchmarking.orgMicroseconds, Fewer Is BetterLiteRT 2024-10-15Model: Inception V48c DDR5-600012c DDR5-60009K18K27K36K45KSE +/- 438.48, N = 6SE +/- 38.95, N = 343959.944088.3

Memcached

Memcached is a high performance, distributed memory object caching system. This Memcached test profiles makes use of memtier_benchmark for excuting this CPU/memory-focused server benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/sec, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:58c DDR5-600012c DDR5-6000800K1600K2400K3200K4000KSE +/- 3725.22, N = 3SE +/- 16589.87, N = 33711825.753701189.231. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.2Blend File: Barbershop - Compute: CPU-Only12c DDR5-60008c DDR5-6000306090120150SE +/- 0.11, N = 3SE +/- 0.12, N = 3120.83121.15

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.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Face Detection Retail FP16-INT8 - Device: CPU12c DDR5-60008c DDR5-60000.89331.78662.67993.57324.4665SE +/- 0.01, N = 3SE +/- 0.00, N = 33.963.97MIN: 1.68 / MAX: 20.25MIN: 1.61 / MAX: 20.641. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

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.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: Danish Mood - Acceleration: CPU8c DDR5-600012c DDR5-60003691215SE +/- 0.11, N = 15SE +/- 0.09, N = 1512.0612.03MIN: 5.76 / MAX: 14.38MIN: 5.72 / MAX: 14.2

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.2Blend File: Classroom - Compute: CPU-Only12c DDR5-60008c DDR5-6000816243240SE +/- 0.04, N = 3SE +/- 0.08, N = 335.4335.51

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.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 128 - Buffer Length: 256 - Filter Length: 5128c DDR5-600012c DDR5-6000400M800M1200M1600M2000MSE +/- 1072898.46, N = 3SE +/- 6251666.44, N = 3195633333319521000001. (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.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Handwritten English Recognition FP16-INT8 - Device: CPU12c DDR5-60008c DDR5-60008001600240032004000SE +/- 4.91, N = 3SE +/- 1.42, N = 33551.353543.791. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Build2

OpenBenchmarking.orgSeconds, Fewer Is BetterBuild2 0.17Time To Compile12c DDR5-60008c DDR5-60001224364860SE +/- 0.01, N = 3SE +/- 0.16, N = 353.4453.55

m-queens

A solver for the N-queens problem with multi-threading support via the OpenMP library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterm-queens 1.2Time To Solve12c DDR5-60008c DDR5-6000246810SE +/- 0.047, N = 3SE +/- 0.010, N = 37.2147.2291. (CXX) g++ options: -fopenmp -O2 -march=native

John The Ripper

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

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: bcrypt12c DDR5-60008c DDR5-600050K100K150K200K250KSE +/- 33.49, N = 3SE +/- 505.86, N = 32375422370551. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.3Encoder Mode: Preset 13 - Input: Beauty 4K 10-bit12c DDR5-60008c DDR5-600048121620SE +/- 0.01, N = 3SE +/- 0.00, N = 317.1517.111. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

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.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Handwritten English Recognition FP16-INT8 - Device: CPU12c DDR5-60008c DDR5-6000612182430SE +/- 0.04, N = 3SE +/- 0.01, N = 327.0127.06MIN: 16.35 / MAX: 47.96MIN: 16.36 / MAX: 45.561. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

PostgreSQL

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 17Scaling Factor: 1000 - Clients: 800 - Mode: Read Only12c DDR5-60008c DDR5-6000700K1400K2100K2800K3500KSE +/- 1685.88, N = 3SE +/- 8342.05, N = 3315503631492771. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Medium8c DDR5-600012c DDR5-60001020304050SE +/- 0.09, N = 3SE +/- 0.06, N = 345.3845.301. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.2Blend File: Pabellon Barcelona - Compute: CPU-Only8c DDR5-600012c DDR5-6000918273645SE +/- 0.15, N = 3SE +/- 0.14, N = 341.1841.25

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.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: Orange Juice - Acceleration: CPU12c DDR5-60008c DDR5-6000612182430SE +/- 0.09, N = 3SE +/- 0.09, N = 324.1924.15MIN: 21.48 / MAX: 33.13MIN: 21.4 / MAX: 32.92

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.3Encoder Mode: Preset 5 - Input: Beauty 4K 10-bit12c DDR5-60008c DDR5-6000246810SE +/- 0.039, N = 3SE +/- 0.026, N = 37.6997.6871. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

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.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Weld Porosity Detection FP16-INT8 - Device: CPU8c DDR5-600012c DDR5-6000246810SE +/- 0.00, N = 3SE +/- 0.01, N = 36.706.71MIN: 2.45 / MAX: 24MIN: 2.58 / MAX: 23.111. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Weld Porosity Detection FP16-INT8 - Device: CPU8c DDR5-600012c DDR5-60003K6K9K12K15KSE +/- 1.42, N = 3SE +/- 11.52, N = 314026.3614006.151. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

oneDNN

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.6Harness: Deconvolution Batch shapes_3d - Engine: CPU8c DDR5-600012c DDR5-60000.16120.32240.48360.64480.806SE +/- 0.001800, N = 3SE +/- 0.002304, N = 30.7153630.716373MIN: 0.62MIN: 0.621. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -fcf-protection=full -pie -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.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Person Re-Identification Retail FP16 - Device: CPU12c DDR5-60008c DDR5-60002K4K6K8K10KSE +/- 5.49, N = 3SE +/- 7.25, N = 310462.5610448.921. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

oneDNN

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.6Harness: Recurrent Neural Network Inference - Engine: CPU12c DDR5-60008c DDR5-600060120180240300SE +/- 0.81, N = 3SE +/- 0.67, N = 3277.64278.00MIN: 269.94MIN: 271.731. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -fcf-protection=full -pie -ldl

ASTC Encoder

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 5.0Preset: Thorough12c DDR5-60008c DDR5-600020406080100SE +/- 0.02, N = 3SE +/- 0.04, N = 382.2882.191. (CXX) g++ options: -O3 -flto -pthread

7-Zip Compression

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 24.05Test: Decompression Rating12c DDR5-60008c DDR5-6000130K260K390K520K650KSE +/- 169.10, N = 3SE +/- 340.29, N = 36272156265531. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

ASTC Encoder

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 5.0Preset: Medium12c DDR5-60008c DDR5-6000120240360480600SE +/- 0.76, N = 3SE +/- 0.83, N = 3568.83568.341. (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.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Face Detection Retail FP16-INT8 - Device: CPU8c DDR5-600012c DDR5-60005K10K15K20K25KSE +/- 11.10, N = 3SE +/- 29.34, N = 323443.8623427.501. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Slow12c DDR5-60008c DDR5-60001020304050SE +/- 0.08, N = 3SE +/- 0.11, N = 344.7944.761. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

XNNPACK

OpenBenchmarking.orgus, Fewer Is BetterXNNPACK b7b048Model: FP32MobileNetV28c DDR5-600012c DDR5-60002K4K6K8K10KSE +/- 46.71, N = 3SE +/- 73.61, N = 3923692421. (CXX) g++ options: -O3 -lrt -lm

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.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 192 - Buffer Length: 256 - Filter Length: 5712c DDR5-60008c DDR5-60001300M2600M3900M5200M6500MSE +/- 9761830.43, N = 3SE +/- 8738103.02, N = 3595860000059554333331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

LiteRT

OpenBenchmarking.orgMicroseconds, Fewer Is BetterLiteRT 2024-10-15Model: SqueezeNet8c DDR5-600012c DDR5-600015003000450060007500SE +/- 27.86, N = 3SE +/- 17.70, N = 37048.397051.59

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.

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.1.1Model: Homogeneous Halfspace12c DDR5-60008c DDR5-6000246810SE +/- 0.106640607, N = 3SE +/- 0.043840287, N = 38.1948624988.1980058371. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -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

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.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Face Detection FP16-INT8 - Device: CPU12c DDR5-60008c DDR5-600070140210280350SE +/- 0.62, N = 3SE +/- 0.26, N = 3330.01330.13MIN: 242.42 / MAX: 353.46MIN: 257.44 / MAX: 357.541. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -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.

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: HMAC-SHA51212c DDR5-60008c DDR5-600090M180M270M360M450MSE +/- 2811379.81, N = 3SE +/- 211622.57, N = 34134743334133263331. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: Blowfish12c DDR5-60008c DDR5-600050K100K150K200K250KSE +/- 16.90, N = 3SE +/- 16.64, N = 32375152374581. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2

ASTC Encoder

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 5.0Preset: Very Thorough8c DDR5-600012c DDR5-60003691215SE +/- 0.00, N = 3SE +/- 0.00, N = 311.6311.631. (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.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Vehicle Detection FP16-INT8 - Device: CPU8c DDR5-600012c DDR5-60002K4K6K8K10KSE +/- 5.48, N = 3SE +/- 10.65, N = 38240.988240.401. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

ASTC Encoder

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 5.0Preset: Exhaustive8c DDR5-600012c DDR5-6000246810SE +/- 0.0004, N = 3SE +/- 0.0018, N = 37.13397.13341. (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.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Face Detection FP16-INT8 - Device: CPU8c DDR5-600012c DDR5-6000306090120150SE +/- 0.12, N = 3SE +/- 0.26, N = 3145.11145.101. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

srsRAN Project

srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.10.1-20240325Test: PUSCH Processor Benchmark, Throughput Total8c DDR5-600012c DDR5-600013002600390052006500SE +/- 0.12, N = 3SE +/- 0.38, N = 36238.16237.7MIN: 3819.5 / MAX: 6238.3MIN: 3818.9 / MAX: 6238.41. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -mavx512f -mavx512cd -mavx512bw -mavx512dq -fno-trapping-math -fno-math-errno -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.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPU8c DDR5-600012c DDR5-60000.07430.14860.22290.29720.3715SE +/- 0.00, N = 3SE +/- 0.00, N = 30.330.33MIN: 0.13 / MAX: 24.63MIN: 0.12 / MAX: 23.431. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Person Re-Identification Retail FP16 - Device: CPU8c DDR5-600012c DDR5-60001.0262.0523.0784.1045.13SE +/- 0.01, N = 3SE +/- 0.00, N = 34.564.56MIN: 2.23 / MAX: 20.98MIN: 2.45 / MAX: 18.951. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Vehicle Detection FP16-INT8 - Device: CPU8c DDR5-600012c DDR5-60001.3052.613.9155.226.525SE +/- 0.00, N = 3SE +/- 0.01, N = 35.805.80MIN: 2.26 / MAX: 26.37MIN: 2.41 / MAX: 27.541. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

PostgreSQL

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 17Scaling Factor: 1000 - Clients: 800 - Mode: Read Only - Average Latency8c DDR5-600012c DDR5-60000.05720.11440.17160.22880.286SE +/- 0.001, N = 3SE +/- 0.000, N = 30.2540.2541. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm

ONNX Runtime

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: CaffeNet 12-int8 - Device: CPU - Executor: Standard12c DDR5-60008c DDR5-60000.23570.47140.70710.94281.1785SE +/- 0.010778, N = 3SE +/- 0.018573, N = 150.9595311.0475051. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: CaffeNet 12-int8 - Device: CPU - Executor: Standard12c DDR5-60008c DDR5-60002004006008001000SE +/- 11.56, N = 3SE +/- 17.42, N = 151041.95958.581. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: CaffeNet 12-int8 - Device: CPU - Executor: Standard12c DDR5-60008c DDR5-60000.23570.47140.70710.94281.1785SE +/- 0.010778, N = 3SE +/- 0.018573, N = 150.9595311.0475051. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

LiteRT

OpenBenchmarking.orgMicroseconds, Fewer Is BetterLiteRT 2024-10-15Model: Mobilenet Quant12c DDR5-60008c DDR5-600013002600390052006500SE +/- 109.66, N = 15SE +/- 155.36, N = 155800.656260.21

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.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: EP.D8c DDR5-600012c DDR5-60003K6K9K12K15KSE +/- 133.00, N = 3SE +/- 614.47, N = 1214658.2513685.971. (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 4.1.6

187 Results Shown

Stream
OpenVINO
Stream
OpenVINO
Stream:
  Add
  Triad
High Performance Conjugate Gradient:
  104 104 104 - 60
  144 144 144 - 60
Algebraic Multi-Grid Benchmark
Xcompact3d Incompact3d
NAS Parallel Benchmarks
OpenRadioss
NAS Parallel Benchmarks:
  MG.C
  FT.C
OpenFOAM
Xcompact3d Incompact3d
NAS Parallel Benchmarks
MBW
NAS Parallel Benchmarks
TensorFlow
MBW
RAMspeed SMP:
  Add - Integer
  Copy - Integer
NAMD
RAMspeed SMP
ClickHouse
RAMspeed SMP
NAS Parallel Benchmarks
ClickHouse
RAMspeed SMP
ClickHouse
OpenVINO:
  Machine Translation EN To DE FP16 - CPU:
    ms
    FPS
NAS Parallel Benchmarks
Graph500
GPAW
Graph500
OpenVINO
Graph500:
  26:
    bfs median_TEPS
    bfs max_TEPS
NAS Parallel Benchmarks
MBW
OpenVINO
MBW
7-Zip Compression
GROMACS
ONNX Runtime
Apache HTTP Server
Whisperfile
OpenVINO:
  Road Segmentation ADAS FP16-INT8 - CPU:
    FPS
    ms
Whisperfile
Apache Cassandra
NAMD
Epoch
ONNX Runtime
LuxCoreRender
XNNPACK
nginx
ASTC Encoder
LiteRT
ONNX Runtime
LiteRT
LuxCoreRender
OpenFOAM
ONNX Runtime
OpenFOAM
uvg266
XNNPACK
SPECFEM3D
OpenVKL
SVT-AV1
Kvazaar
Liquid-DSP
ONNX Runtime
uvg266
Timed Linux Kernel Compilation:
  allmodconfig
  defconfig
VVenC
LiteRT
PostgreSQL
XNNPACK
PostgreSQL
LuxCoreRender
SPECFEM3D
NAS Parallel Benchmarks
ONNX Runtime
XNNPACK
John The Ripper
Kvazaar
XNNPACK
Memcached
uvg266
SPECFEM3D
oneDNN
LiteRT
XNNPACK
WarpX
OpenFOAM
SVT-AV1
RawTherapee
srsRAN Project
LiteRT
Stargate Digital Audio Workstation:
  192000 - 1024
  192000 - 512
oneDNN
Memcached
XNNPACK
Chaos Group V-RAY
OpenVINO
XNNPACK
OpenRadioss
Kvazaar
OpenVINO
oneDNN
SVT-AV1
Liquid-DSP:
  192 - 256 - 512
  64 - 256 - 57
SPECFEM3D
oneDNN
miniBUDE:
  OpenMP - BM2:
    GFInst/s
    Billion Interactions/s
SVT-AV1
Blender
VVenC
oneDNN
Stargate Digital Audio Workstation
Timed Godot Game Engine Compilation
John The Ripper
Blender
OpenVINO
uvg266
WarpX
Stargate Digital Audio Workstation
Liquid-DSP
uvg266
SVT-AV1:
  Preset 3 - Beauty 4K 10-bit
  Preset 8 - Beauty 4K 10-bit
Blender
LiteRT
Memcached
Blender
OpenVINO
LuxCoreRender
Blender
Liquid-DSP
OpenVINO
Build2
m-queens
John The Ripper
SVT-AV1
OpenVINO
PostgreSQL
Kvazaar
Blender
LuxCoreRender
SVT-AV1
OpenVINO:
  Weld Porosity Detection FP16-INT8 - CPU:
    ms
    FPS
oneDNN
OpenVINO
oneDNN
ASTC Encoder
7-Zip Compression
ASTC Encoder
OpenVINO
Kvazaar
XNNPACK
Liquid-DSP
LiteRT
SPECFEM3D
OpenVINO
John The Ripper:
  HMAC-SHA512
  Blowfish
ASTC Encoder
OpenVINO
ASTC Encoder
OpenVINO
srsRAN Project
OpenVINO:
  Age Gender Recognition Retail 0013 FP16-INT8 - CPU
  Person Re-Identification Retail FP16 - CPU
  Vehicle Detection FP16-INT8 - CPU
PostgreSQL
ONNX Runtime:
  CaffeNet 12-int8 - CPU - Standard:
    Inference Time Cost (ms)
    Inferences Per Second
LiteRT
NAS Parallel Benchmarks

8c DDR5-6000

Processor: AMD EPYC 9655P 96-Core @ 2.60GHz (96 Cores / 192 Threads), Motherboard: Supermicro Super Server H13SSL-N v1.01 (3.0 BIOS), Chipset: AMD 1Ah, Memory: 8 x 64GB DDR5-6000MT/s, Disk: 3201GB Micron_7450_MTFDKCB3T2TFS, Graphics: ASPEED, Network: 2 x Broadcom NetXtreme BCM5720 PCIe

OS: Ubuntu 24.10, Kernel: 6.12.0-rc7-phx (x86_64), Desktop: GNOME Shell 47.0, Display Server: X Server, Compiler: GCC 14.2.0, 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,d,fortran,objc,obj-c++,m2,rust --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-14-zdkDXv/gcc-14-14.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-14-zdkDXv/gcc-14-14.2.0/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: 0xb002116
Java Notes: OpenJDK Runtime Environment (build 21.0.5+11-Ubuntu-1ubuntu124.10)
Python Notes: Python 3.12.7
Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: 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 Enhanced / Automatic IBRS; IBPB: conditional; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 15 November 2024 21:45 by user phoronix.

12c DDR5-6000

Processor: AMD EPYC 9655P 96-Core @ 2.60GHz (96 Cores / 192 Threads), Motherboard: Supermicro Super Server H13SSL-N v1.01 (3.0 BIOS), Chipset: AMD 1Ah, Memory: 12 x 64GB DDR5-6000MT/s Micron MTC40F2046S1RC64BDY QSFF, Disk: 3201GB Micron_7450_MTFDKCB3T2TFS, Graphics: ASPEED, Network: 2 x Broadcom NetXtreme BCM5720 PCIe

OS: Ubuntu 24.10, Kernel: 6.12.0-rc7-phx (x86_64), Desktop: GNOME Shell 47.0, Display Server: X Server, Compiler: GCC 14.2.0, 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,d,fortran,objc,obj-c++,m2,rust --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-14-zdkDXv/gcc-14-14.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-14-zdkDXv/gcc-14-14.2.0/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: 0xb002116
Python Notes: Python 3.12.7
Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: 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 Enhanced / Automatic IBRS; IBPB: conditional; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 15 November 2024 11:35 by user phoronix.