gh200

ARMv8 Neoverse-V2 testing with a Pegatron JIMBO P4352 (00022432 BIOS) and NVIDIA GH200 144G HBM3e 143GB on Ubuntu 24.04 via the Phoronix Test Suite.

HTML result view exported from: https://openbenchmarking.org/result/2410122-NE-GH200545357&grt.

gh200ProcessorMotherboardMemoryDiskGraphicsNetworkOSKernelDisplay DriverOpenCLCompilerFile-SystemScreen ResolutionaARMv8 Neoverse-V2 @ 3.47GHz (72 Cores)Pegatron JIMBO P4352 (00022432 BIOS)1 x 480GB LPDDR5-6400MT/s NVIDIA 699-2G530-0236-RC11000GB CT1000T700SSD3NVIDIA GH200 144G HBM3e 143GB2 x Intel X550Ubuntu 24.046.8.0-45-generic-64k (aarch64)NVIDIAOpenCL 3.0 CUDA 12.6.65GCC 13.2.0ext41920x1200OpenBenchmarking.org- Transparent Huge Pages: madvise- --build=aarch64-linux-gnu --disable-libquadmath --disable-libquadmath-support --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-fix-cortex-a53-843419 --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-dIwDw0/gcc-13-13.2.0/debian/tmp-nvptx/usr --enable-plugin --enable-shared --enable-threads=posix --host=aarch64-linux-gnu --program-prefix=aarch64-linux-gnu- --target=aarch64-linux-gnu --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-target-system-zlib=auto --without-cuda-driver -v - Scaling Governor: cppc_cpufreq ondemand (Boost: Disabled)- OpenJDK Runtime Environment (build 21.0.4+7-Ubuntu-1ubuntu224.04)- Python 3.12.3- 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 __user pointer sanitization + spectre_v2: Not affected + srbds: Not affected + tsx_async_abort: Not affected

gh200compress-7zip: Compression Ratingcompress-7zip: Decompression Ratingcompress-7zip: Compression Ratingcompress-7zip: Decompression Ratingblender: BMW27 - CPU-Onlyblender: Classroom - CPU-Onlyblender: Fishy Cat - CPU-Onlyblender: Barbershop - CPU-Onlyblender: Pabellon Barcelona - CPU-Onlybuild2: Time To Compilebyte: Pipebyte: Dhrystone 2byte: System Callbyte: Whetstone Doublec-ray: 4K - 16c-ray: 5K - 16c-ray: 1080p - 16epoch: Coneetcpak: Multi-Threaded - ETC2graphics-magick: Swirlgraphics-magick: Rotategraphics-magick: Sharpengraphics-magick: Enhancedgraphics-magick: Resizinggraphics-magick: Noise-Gaussiangraphics-magick: HWB Color Spacegraphics-magick: Swirlgraphics-magick: Rotategraphics-magick: Sharpengraphics-magick: Enhancedgraphics-magick: Resizinggraphics-magick: Noise-Gaussiangraphics-magick: HWB Color Spacegromacs: water_GMX50_baregromacs: MPI CPU - water_GMX50_barelczero: Eigenmnn: nasnetmnn: mobilenetV3mnn: squeezenetv1.1mnn: resnet-v2-50mnn: SqueezeNetV1.0mnn: MobileNetV2_224mnn: mobilenet-v1-1.0mnn: inception-v3onnx: yolov4 - CPU - Parallelonnx: yolov4 - CPU - Parallelonnx: yolov4 - CPU - Standardonnx: yolov4 - CPU - Standardonnx: ZFNet-512 - CPU - Parallelonnx: ZFNet-512 - CPU - Parallelonnx: ZFNet-512 - CPU - Standardonnx: ZFNet-512 - CPU - Standardonnx: T5 Encoder - CPU - Parallelonnx: T5 Encoder - CPU - Parallelonnx: T5 Encoder - CPU - Standardonnx: T5 Encoder - CPU - Standardonnx: CaffeNet 12-int8 - CPU - Parallelonnx: CaffeNet 12-int8 - CPU - Parallelonnx: CaffeNet 12-int8 - CPU - Standardonnx: CaffeNet 12-int8 - CPU - Standardonnx: fcn-resnet101-11 - CPU - Parallelonnx: fcn-resnet101-11 - CPU - Parallelonnx: fcn-resnet101-11 - CPU - Standardonnx: fcn-resnet101-11 - CPU - Standardonnx: ResNet50 v1-12-int8 - CPU - Parallelonnx: ResNet50 v1-12-int8 - CPU - Parallelonnx: ResNet50 v1-12-int8 - CPU - Standardonnx: ResNet50 v1-12-int8 - CPU - Standardonnx: super-resolution-10 - CPU - Parallelonnx: super-resolution-10 - CPU - Parallelonnx: super-resolution-10 - CPU - Standardonnx: super-resolution-10 - CPU - Standardonnx: ResNet101_DUC_HDC-12 - CPU - Parallelonnx: ResNet101_DUC_HDC-12 - CPU - Parallelonnx: ResNet101_DUC_HDC-12 - CPU - Standardonnx: ResNet101_DUC_HDC-12 - CPU - Standardpovray: Trace Timepyperformance: gopyperformance: chaospyperformance: floatpyperformance: nbodypyperformance: pathlibpyperformance: raytracepyperformance: xml_etreepyperformance: gc_collectpyperformance: json_loadspyperformance: crypto_pyaespyperformance: async_tree_iopyperformance: regex_compilepyperformance: python_startuppyperformance: asyncio_tcp_sslpyperformance: django_templatepyperformance: asyncio_websocketspyperformance: pickle_pure_pythonsimdjson: Kostyasimdjson: TopTweetsimdjson: LargeRandsimdjson: PartialTweetssimdjson: DistinctUserIDstockfish: Chess Benchmarkstockfish: Chess Benchmarkbuild-linux-kernel: defconfigbuild-linux-kernel: allmodconfigbuild-llvm: Ninjabuild-llvm: Unix Makefileswarpx: Uniform Plasmawarpx: Plasma Accelerationx265: Bosphorus 4Kx265: Bosphorus 1080pxnnpack: FP32MobileNetV2xnnpack: FP32MobileNetV3Largexnnpack: FP32MobileNetV3Smallxnnpack: FP16MobileNetV2xnnpack: FP16MobileNetV3Largexnnpack: FP16MobileNetV3Smallxnnpack: QU8MobileNetV2xnnpack: QU8MobileNetV3Largexnnpack: QU8MobileNetV3Smalla39352341881938477542052438.0678.3773.02381.45154.4684.786202565282.24998587529.8145868649.3721978.020.35736.2085.195188.20471.1936052091713512822174306573314113594423016567.1566.0013605.0081.1341.82411.3353.3961.5021.79313.6935.80711172.2255.14685194.36445.231422.1451210.1174.75865109.2419.15193390.0672.56003321.2373.111581262.140.7916240.5829021715.640.4629432160.11161.3326.19726317.8683.1452318.648753.6252159.0686.285530.3601962776.330.3032223298.197.78698.247.456.864.515.521745.81.0817.554.874882.318.71.4926.35102053.114.141.154.064.165849675316842876366.710285.132175.030276.92616.9020207620.380538218.8112.619671426945840122688194514841083OpenBenchmarking.org

7-Zip Compression

Test: Compression Rating

OpenBenchmarking.orgMIPS, More Is Better7-Zip CompressionTest: Compression Ratinga80K160K240K320K400KSE +/- 3097.20, N = 33935231. 7-Zip 23.01 (arm64) : Copyright (c) 1999-2023 Igor Pavlov : 2023-06-20

7-Zip Compression

Test: Decompression Rating

OpenBenchmarking.orgMIPS, More Is Better7-Zip CompressionTest: Decompression Ratinga90K180K270K360K450KSE +/- 507.84, N = 34188191. 7-Zip 23.01 (arm64) : Copyright (c) 1999-2023 Igor Pavlov : 2023-06-20

7-Zip Compression

Test: Compression Rating

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 24.05Test: Compression Ratinga80K160K240K320K400KSE +/- 4213.31, N = 33847751. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

7-Zip Compression

Test: Decompression Rating

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 24.05Test: Decompression Ratinga90K180K270K360K450KSE +/- 944.71, N = 34205241. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

Blender

Blend File: BMW27 - Compute: CPU-Only

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.0.2Blend File: BMW27 - Compute: CPU-Onlya918273645SE +/- 0.04, N = 338.06

Blender

Blend File: Classroom - Compute: CPU-Only

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.0.2Blend File: Classroom - Compute: CPU-Onlya20406080100SE +/- 0.08, N = 378.37

Blender

Blend File: Fishy Cat - Compute: CPU-Only

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.0.2Blend File: Fishy Cat - Compute: CPU-Onlya1632486480SE +/- 0.44, N = 373.02

Blender

Blend File: Barbershop - Compute: CPU-Only

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.0.2Blend File: Barbershop - Compute: CPU-Onlya80160240320400SE +/- 0.54, N = 3381.45

Blender

Blend File: Pabellon Barcelona - Compute: CPU-Only

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.0.2Blend File: Pabellon Barcelona - Compute: CPU-Onlya306090120150SE +/- 0.37, N = 3154.46

Build2

Time To Compile

OpenBenchmarking.orgSeconds, Fewer Is BetterBuild2 0.17Time To Compilea20406080100SE +/- 0.22, N = 384.79

BYTE Unix Benchmark

Computational Test: Pipe

OpenBenchmarking.orgLPS, More Is BetterBYTE Unix Benchmark 5.1.3-gitComputational Test: Pipea40M80M120M160M200MSE +/- 32087.94, N = 3202565282.21. (CC) gcc options: -pedantic -O3 -ffast-math -march=native -mtune=native -lm

BYTE Unix Benchmark

Computational Test: Dhrystone 2

OpenBenchmarking.orgLPS, More Is BetterBYTE Unix Benchmark 5.1.3-gitComputational Test: Dhrystone 2a1100M2200M3300M4400M5500MSE +/- 2591819.88, N = 34998587529.81. (CC) gcc options: -pedantic -O3 -ffast-math -march=native -mtune=native -lm

BYTE Unix Benchmark

Computational Test: System Call

OpenBenchmarking.orgLPS, More Is BetterBYTE Unix Benchmark 5.1.3-gitComputational Test: System Calla30M60M90M120M150MSE +/- 15202.21, N = 3145868649.31. (CC) gcc options: -pedantic -O3 -ffast-math -march=native -mtune=native -lm

BYTE Unix Benchmark

Computational Test: Whetstone Double

OpenBenchmarking.orgMWIPS, More Is BetterBYTE Unix Benchmark 5.1.3-gitComputational Test: Whetstone Doublea150K300K450K600K750KSE +/- 19.25, N = 3721978.01. (CC) gcc options: -pedantic -O3 -ffast-math -march=native -mtune=native -lm

C-Ray

Resolution: 4K - Rays Per Pixel: 16

OpenBenchmarking.orgSeconds, Fewer Is BetterC-Ray 2.0Resolution: 4K - Rays Per Pixel: 16a510152025SE +/- 0.00, N = 320.361. (CC) gcc options: -lpthread -lm

C-Ray

Resolution: 5K - Rays Per Pixel: 16

OpenBenchmarking.orgSeconds, Fewer Is BetterC-Ray 2.0Resolution: 5K - Rays Per Pixel: 16a816243240SE +/- 0.02, N = 336.211. (CC) gcc options: -lpthread -lm

C-Ray

Resolution: 1080p - Rays Per Pixel: 16

OpenBenchmarking.orgSeconds, Fewer Is BetterC-Ray 2.0Resolution: 1080p - Rays Per Pixel: 16a1.16892.33783.50674.67565.8445SE +/- 0.003, N = 35.1951. (CC) gcc options: -lpthread -lm

Epoch

Epoch3D Deck: Cone

OpenBenchmarking.orgSeconds, Fewer Is BetterEpoch 4.19.4Epoch3D Deck: Conea4080120160200SE +/- 2.18, N = 4188.201. (F9X) gfortran options: -O3 -std=f2003 -Jobj -lsdf -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

Etcpak

Benchmark: Multi-Threaded - Configuration: ETC2

OpenBenchmarking.orgMpx/s, More Is BetterEtcpak 2.0Benchmark: Multi-Threaded - Configuration: ETC2a100200300400500SE +/- 2.47, N = 3471.191. (CXX) g++ options: -flto -pthread

GraphicsMagick

Operation: Swirl

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagickOperation: Swirla130260390520650SE +/- 5.51, N = 36051. GraphicsMagick 1.3.42 2023-09-23 Q16 http://www.GraphicsMagick.org/

GraphicsMagick

Operation: Rotate

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagickOperation: Rotatea50100150200250SE +/- 0.88, N = 32091. GraphicsMagick 1.3.42 2023-09-23 Q16 http://www.GraphicsMagick.org/

GraphicsMagick

Operation: Sharpen

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagickOperation: Sharpena4080120160200SE +/- 0.33, N = 31711. GraphicsMagick 1.3.42 2023-09-23 Q16 http://www.GraphicsMagick.org/

GraphicsMagick

Operation: Enhanced

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagickOperation: Enhanceda80160240320400SE +/- 0.33, N = 33511. GraphicsMagick 1.3.42 2023-09-23 Q16 http://www.GraphicsMagick.org/

GraphicsMagick

Operation: Resizing

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagickOperation: Resizinga60120180240300SE +/- 1.15, N = 32821. GraphicsMagick 1.3.42 2023-09-23 Q16 http://www.GraphicsMagick.org/

GraphicsMagick

Operation: Noise-Gaussian

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagickOperation: Noise-Gaussiana50100150200250SE +/- 1.73, N = 32171. GraphicsMagick 1.3.42 2023-09-23 Q16 http://www.GraphicsMagick.org/

GraphicsMagick

Operation: HWB Color Space

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagickOperation: HWB Color Spacea90180270360450SE +/- 0.67, N = 34301. GraphicsMagick 1.3.42 2023-09-23 Q16 http://www.GraphicsMagick.org/

GraphicsMagick

Operation: Swirl

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: Swirla140280420560700SE +/- 4.26, N = 36571. (CC) gcc options: -fopenmp -O2 -ljpeg -lSM -lICE -lX11 -lz -lm -lpthread -lgomp

GraphicsMagick

Operation: Rotate

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: Rotatea70140210280350SE +/- 4.26, N = 33311. (CC) gcc options: -fopenmp -O2 -ljpeg -lSM -lICE -lX11 -lz -lm -lpthread -lgomp

GraphicsMagick

Operation: Sharpen

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: Sharpena90180270360450SE +/- 0.58, N = 34111. (CC) gcc options: -fopenmp -O2 -ljpeg -lSM -lICE -lX11 -lz -lm -lpthread -lgomp

GraphicsMagick

Operation: Enhanced

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: Enhanceda80160240320400SE +/- 0.67, N = 33591. (CC) gcc options: -fopenmp -O2 -ljpeg -lSM -lICE -lX11 -lz -lm -lpthread -lgomp

GraphicsMagick

Operation: Resizing

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: Resizinga100200300400500SE +/- 6.06, N = 34421. (CC) gcc options: -fopenmp -O2 -ljpeg -lSM -lICE -lX11 -lz -lm -lpthread -lgomp

GraphicsMagick

Operation: Noise-Gaussian

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: Noise-Gaussiana70140210280350SE +/- 2.18, N = 153011. (CC) gcc options: -fopenmp -O2 -ljpeg -lSM -lICE -lX11 -lz -lm -lpthread -lgomp

GraphicsMagick

Operation: HWB Color Space

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: HWB Color Spacea140280420560700SE +/- 8.82, N = 36561. (CC) gcc options: -fopenmp -O2 -ljpeg -lSM -lICE -lX11 -lz -lm -lpthread -lgomp

GROMACS

Input: water_GMX50_bare

OpenBenchmarking.orgNs Per Day, More Is BetterGROMACSInput: water_GMX50_barea246810SE +/- 0.004, N = 37.1561. GROMACS version: 2023.3-Ubuntu_2023.3_1ubuntu3

GROMACS

Implementation: MPI CPU - Input: water_GMX50_bare

OpenBenchmarking.orgNs Per Day, More Is BetterGROMACS 2024Implementation: MPI CPU - Input: water_GMX50_barea246810SE +/- 0.003, N = 36.0011. (CXX) g++ options: -O3 -lm

LeelaChessZero

Backend: Eigen

OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.31.1Backend: Eigena80160240320400SE +/- 4.26, N = 33601. (CXX) g++ options: -flto -pthread

Mobile Neural Network

Model: nasnet

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.9.b11b7037dModel: nasneta1.12682.25363.38044.50725.634SE +/- 0.036, N = 35.008MIN: 4.49 / MAX: 27.911. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -pthread -ldl

Mobile Neural Network

Model: mobilenetV3

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.9.b11b7037dModel: mobilenetV3a0.25520.51040.76561.02081.276SE +/- 0.009, N = 31.134MIN: 0.69 / MAX: 11.141. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -pthread -ldl

Mobile Neural Network

Model: squeezenetv1.1

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.9.b11b7037dModel: squeezenetv1.1a0.41040.82081.23121.64162.052SE +/- 0.044, N = 31.824MIN: 1.17 / MAX: 20.371. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -pthread -ldl

Mobile Neural Network

Model: resnet-v2-50

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.9.b11b7037dModel: resnet-v2-50a3691215SE +/- 0.10, N = 311.34MIN: 8.54 / MAX: 42.161. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -pthread -ldl

Mobile Neural Network

Model: SqueezeNetV1.0

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.9.b11b7037dModel: SqueezeNetV1.0a0.76411.52822.29233.05643.8205SE +/- 0.027, N = 33.396MIN: 2.14 / MAX: 29.881. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -pthread -ldl

Mobile Neural Network

Model: MobileNetV2_224

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.9.b11b7037dModel: MobileNetV2_224a0.3380.6761.0141.3521.69SE +/- 0.019, N = 31.502MIN: 1.12 / MAX: 13.521. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -pthread -ldl

Mobile Neural Network

Model: mobilenet-v1-1.0

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.9.b11b7037dModel: mobilenet-v1-1.0a0.40340.80681.21021.61362.017SE +/- 0.005, N = 31.793MIN: 1.34 / MAX: 22.051. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -pthread -ldl

Mobile Neural Network

Model: inception-v3

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.9.b11b7037dModel: inception-v3a48121620SE +/- 0.02, N = 313.69MIN: 11.51 / MAX: 42.341. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -pthread -ldl

ONNX Runtime

Model: yolov4 - Device: CPU - Executor: Parallel

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: yolov4 - Device: CPU - Executor: Parallela1.30662.61323.91985.22646.533SE +/- 0.04903, N = 35.807111. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: yolov4 - Device: CPU - Executor: Parallel

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: yolov4 - Device: CPU - Executor: Parallela4080120160200SE +/- 1.46, N = 3172.231. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: yolov4 - Device: CPU - Executor: Standard

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: yolov4 - Device: CPU - Executor: Standarda1.1582.3163.4744.6325.79SE +/- 0.07121, N = 35.146851. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: yolov4 - Device: CPU - Executor: Standard

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: yolov4 - Device: CPU - Executor: Standarda4080120160200SE +/- 2.71, N = 3194.361. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: ZFNet-512 - Device: CPU - Executor: Parallel

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: ZFNet-512 - Device: CPU - Executor: Parallela1020304050SE +/- 0.50, N = 1545.231. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: ZFNet-512 - Device: CPU - Executor: Parallel

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: ZFNet-512 - Device: CPU - Executor: Parallela510152025SE +/- 0.25, N = 1522.151. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: ZFNet-512 - Device: CPU - Executor: Standard

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: ZFNet-512 - Device: CPU - Executor: Standarda50100150200250SE +/- 2.64, N = 3210.121. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: ZFNet-512 - Device: CPU - Executor: Standard

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: ZFNet-512 - Device: CPU - Executor: Standarda1.07072.14143.21214.28285.3535SE +/- 0.05942, N = 34.758651. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: T5 Encoder - Device: CPU - Executor: Parallel

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: T5 Encoder - Device: CPU - Executor: Parallela20406080100SE +/- 0.76, N = 3109.241. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: T5 Encoder - Device: CPU - Executor: Parallel

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: T5 Encoder - Device: CPU - Executor: Parallela3691215SE +/- 0.06344, N = 39.151931. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: T5 Encoder - Device: CPU - Executor: Standard

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: T5 Encoder - Device: CPU - Executor: Standarda80160240320400SE +/- 2.61, N = 3390.071. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: T5 Encoder - Device: CPU - Executor: Standard

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: T5 Encoder - Device: CPU - Executor: Standarda0.5761.1521.7282.3042.88SE +/- 0.01750, N = 32.560031. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: CaffeNet 12-int8 - Device: CPU - Executor: Parallel

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: CaffeNet 12-int8 - Device: CPU - Executor: Parallela70140210280350SE +/- 0.27, N = 3321.241. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: CaffeNet 12-int8 - Device: CPU - Executor: Parallel

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: CaffeNet 12-int8 - Device: CPU - Executor: Parallela0.70011.40022.10032.80043.5005SE +/- 0.00260, N = 33.111581. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: CaffeNet 12-int8 - Device: CPU - Executor: Standard

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: CaffeNet 12-int8 - Device: CPU - Executor: Standarda30060090012001500SE +/- 2.68, N = 31262.141. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: CaffeNet 12-int8 - Device: CPU - Executor: Standard

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: CaffeNet 12-int8 - Device: CPU - Executor: Standarda0.17810.35620.53430.71240.8905SE +/- 0.001678, N = 30.7916241. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: fcn-resnet101-11 - Device: CPU - Executor: Parallel

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: fcn-resnet101-11 - Device: CPU - Executor: Parallela0.13120.26240.39360.52480.656SE +/- 0.002945, N = 30.5829021. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: fcn-resnet101-11 - Device: CPU - Executor: Parallel

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: fcn-resnet101-11 - Device: CPU - Executor: Parallela400800120016002000SE +/- 8.64, N = 31715.641. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: fcn-resnet101-11 - Device: CPU - Executor: Standard

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: fcn-resnet101-11 - Device: CPU - Executor: Standarda0.10420.20840.31260.41680.521SE +/- 0.001019, N = 30.4629431. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: fcn-resnet101-11 - Device: CPU - Executor: Standard

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: fcn-resnet101-11 - Device: CPU - Executor: Standarda5001000150020002500SE +/- 4.77, N = 32160.111. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: ResNet50 v1-12-int8 - Device: CPU - Executor: Parallel

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: ResNet50 v1-12-int8 - Device: CPU - Executor: Parallela4080120160200SE +/- 0.70, N = 3161.331. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: ResNet50 v1-12-int8 - Device: CPU - Executor: Parallel

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: ResNet50 v1-12-int8 - Device: CPU - Executor: Parallela246810SE +/- 0.02687, N = 36.197261. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: ResNet50 v1-12-int8 - Device: CPU - Executor: Standard

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: ResNet50 v1-12-int8 - Device: CPU - Executor: Standarda70140210280350SE +/- 0.61, N = 3317.871. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: ResNet50 v1-12-int8 - Device: CPU - Executor: Standard

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: ResNet50 v1-12-int8 - Device: CPU - Executor: Standarda0.70771.41542.12312.83083.5385SE +/- 0.00610, N = 33.145231. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: super-resolution-10 - Device: CPU - Executor: Parallel

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: super-resolution-10 - Device: CPU - Executor: Parallela510152025SE +/- 0.10, N = 318.651. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: super-resolution-10 - Device: CPU - Executor: Parallel

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: super-resolution-10 - Device: CPU - Executor: Parallela1224364860SE +/- 0.29, N = 353.631. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: super-resolution-10 - Device: CPU - Executor: Standard

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: super-resolution-10 - Device: CPU - Executor: Standarda4080120160200SE +/- 1.12, N = 3159.071. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: super-resolution-10 - Device: CPU - Executor: Standard

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: super-resolution-10 - Device: CPU - Executor: Standarda246810SE +/- 0.04448, N = 36.285531. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: ResNet101_DUC_HDC-12 - Device: CPU - Executor: Parallel

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: ResNet101_DUC_HDC-12 - Device: CPU - Executor: Parallela0.0810.1620.2430.3240.405SE +/- 0.001260, N = 30.3601961. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: ResNet101_DUC_HDC-12 - Device: CPU - Executor: Parallel

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: ResNet101_DUC_HDC-12 - Device: CPU - Executor: Parallela6001200180024003000SE +/- 9.74, N = 32776.331. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: ResNet101_DUC_HDC-12 - Device: CPU - Executor: Standard

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.19Model: ResNet101_DUC_HDC-12 - Device: CPU - Executor: Standarda0.06820.13640.20460.27280.341SE +/- 0.001997, N = 30.3032221. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: ResNet101_DUC_HDC-12 - Device: CPU - Executor: Standard

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.19Model: ResNet101_DUC_HDC-12 - Device: CPU - Executor: Standarda7001400210028003500SE +/- 21.77, N = 33298.191. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

POV-Ray

Trace Time

OpenBenchmarking.orgSeconds, Fewer Is BetterPOV-RayTrace Timea246810SE +/- 0.061, N = 37.7861. POV-Ray 3.7.0.10.unofficial

PyPerformance

Benchmark: go

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.11Benchmark: goa20406080100SE +/- 0.07, N = 398.2

PyPerformance

Benchmark: chaos

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.11Benchmark: chaosa1122334455SE +/- 0.06, N = 347.4

PyPerformance

Benchmark: float

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.11Benchmark: floata1326395265SE +/- 0.03, N = 356.8

PyPerformance

Benchmark: nbody

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.11Benchmark: nbodya1428425670SE +/- 0.09, N = 364.5

PyPerformance

Benchmark: pathlib

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.11Benchmark: pathliba48121620SE +/- 0.03, N = 315.5

PyPerformance

Benchmark: raytrace

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.11Benchmark: raytracea50100150200250SE +/- 0.33, N = 3217

PyPerformance

Benchmark: xml_etree

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.11Benchmark: xml_etreea1020304050SE +/- 0.03, N = 345.8

PyPerformance

Benchmark: gc_collect

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.11Benchmark: gc_collecta0.2430.4860.7290.9721.215SE +/- 0.01, N = 151.08

PyPerformance

Benchmark: json_loads

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.11Benchmark: json_loadsa48121620SE +/- 0.06, N = 317.5

PyPerformance

Benchmark: crypto_pyaes

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.11Benchmark: crypto_pyaesa1224364860SE +/- 0.03, N = 354.8

PyPerformance

Benchmark: async_tree_io

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.11Benchmark: async_tree_ioa160320480640800SE +/- 2.96, N = 3748

PyPerformance

Benchmark: regex_compile

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.11Benchmark: regex_compilea20406080100SE +/- 0.12, N = 382.3

PyPerformance

Benchmark: python_startup

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.11Benchmark: python_startupa510152025SE +/- 0.06, N = 318.7

PyPerformance

Benchmark: asyncio_tcp_ssl

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.11Benchmark: asyncio_tcp_ssla0.33530.67061.00591.34121.6765SE +/- 0.00, N = 31.49

PyPerformance

Benchmark: django_template

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.11Benchmark: django_templatea612182430SE +/- 0.12, N = 326.3

PyPerformance

Benchmark: asyncio_websockets

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.11Benchmark: asyncio_websocketsa110220330440550SE +/- 0.33, N = 3510

PyPerformance

Benchmark: pickle_pure_python

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.11Benchmark: pickle_pure_pythona4080120160200SE +/- 0.33, N = 3205

simdjson

Throughput Test: Kostya

OpenBenchmarking.orgGB/s, More Is Bettersimdjson 3.10Throughput Test: Kostyaa0.69981.39962.09942.79923.499SE +/- 0.01, N = 33.111. (CXX) g++ options: -O3 -lrt

simdjson

Throughput Test: TopTweet

OpenBenchmarking.orgGB/s, More Is Bettersimdjson 3.10Throughput Test: TopTweeta0.93151.8632.79453.7264.6575SE +/- 0.01, N = 34.141. (CXX) g++ options: -O3 -lrt

simdjson

Throughput Test: LargeRandom

OpenBenchmarking.orgGB/s, More Is Bettersimdjson 3.10Throughput Test: LargeRandoma0.25880.51760.77641.03521.294SE +/- 0.00, N = 31.151. (CXX) g++ options: -O3 -lrt

simdjson

Throughput Test: PartialTweets

OpenBenchmarking.orgGB/s, More Is Bettersimdjson 3.10Throughput Test: PartialTweetsa0.91351.8272.74053.6544.5675SE +/- 0.00, N = 34.061. (CXX) g++ options: -O3 -lrt

simdjson

Throughput Test: DistinctUserID

OpenBenchmarking.orgGB/s, More Is Bettersimdjson 3.10Throughput Test: DistinctUserIDa0.9361.8722.8083.7444.68SE +/- 0.00, N = 34.161. (CXX) g++ options: -O3 -lrt

Stockfish

Chess Benchmark

OpenBenchmarking.orgNodes Per Second, More Is BetterStockfishChess Benchmarka13M26M39M52M65MSE +/- 959000.15, N = 15584967531. Stockfish 16 by the Stockfish developers (see AUTHORS file)

Stockfish

Chess Benchmark

OpenBenchmarking.orgNodes Per Second, More Is BetterStockfish 17Chess Benchmarka40M80M120M160M200MSE +/- 6156005.01, N = 151684287631. (CXX) g++ options: -lgcov -lpthread -fno-exceptions -std=c++17 -fno-peel-loops -fno-tracer -pedantic -O3 -funroll-loops -flto -flto-partition=one -flto=jobserver

Timed Linux Kernel Compilation

Build: defconfig

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.8Build: defconfiga1530456075SE +/- 0.55, N = 1366.71

Timed Linux Kernel Compilation

Build: allmodconfig

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.8Build: allmodconfiga60120180240300SE +/- 2.59, N = 3285.13

Timed LLVM Compilation

Build System: Ninja

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 16.0Build System: Ninjaa4080120160200SE +/- 1.19, N = 3175.03

Timed LLVM Compilation

Build System: Unix Makefiles

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 16.0Build System: Unix Makefilesa60120180240300SE +/- 0.32, N = 3276.93

WarpX

Input: Uniform Plasma

OpenBenchmarking.orgSeconds, Fewer Is BetterWarpX 24.10Input: Uniform Plasmaa48121620SE +/- 0.18, N = 316.901. (CXX) g++ options: -O3

WarpX

Input: Plasma Acceleration

OpenBenchmarking.orgSeconds, Fewer Is BetterWarpX 24.10Input: Plasma Accelerationa510152025SE +/- 0.03, N = 320.381. (CXX) g++ options: -O3

x265

Video Input: Bosphorus 4K

OpenBenchmarking.orgFrames Per Second, More Is Betterx265Video Input: Bosphorus 4Ka246810SE +/- 0.03, N = 38.811. x265 [info]: HEVC encoder version 3.5+1-f0c1022b6

x265

Video Input: Bosphorus 1080p

OpenBenchmarking.orgFrames Per Second, More Is Betterx265Video Input: Bosphorus 1080pa3691215SE +/- 0.18, N = 312.611. x265 [info]: HEVC encoder version 3.5+1-f0c1022b6

XNNPACK

Model: FP32MobileNetV2

OpenBenchmarking.orgus, Fewer Is BetterXNNPACK 2cd86bModel: FP32MobileNetV2a2004006008001000SE +/- 8.41, N = 39671. (CXX) g++ options: -O3 -lrt -lm

XNNPACK

Model: FP32MobileNetV3Large

OpenBenchmarking.orgus, Fewer Is BetterXNNPACK 2cd86bModel: FP32MobileNetV3Largea30060090012001500SE +/- 6.51, N = 314261. (CXX) g++ options: -O3 -lrt -lm

XNNPACK

Model: FP32MobileNetV3Small

OpenBenchmarking.orgus, Fewer Is BetterXNNPACK 2cd86bModel: FP32MobileNetV3Smalla2004006008001000SE +/- 16.38, N = 39451. (CXX) g++ options: -O3 -lrt -lm

XNNPACK

Model: FP16MobileNetV2

OpenBenchmarking.orgus, Fewer Is BetterXNNPACK 2cd86bModel: FP16MobileNetV2a2004006008001000SE +/- 15.62, N = 38401. (CXX) g++ options: -O3 -lrt -lm

XNNPACK

Model: FP16MobileNetV3Large

OpenBenchmarking.orgus, Fewer Is BetterXNNPACK 2cd86bModel: FP16MobileNetV3Largea30060090012001500SE +/- 21.31, N = 312261. (CXX) g++ options: -O3 -lrt -lm

XNNPACK

Model: FP16MobileNetV3Small

OpenBenchmarking.orgus, Fewer Is BetterXNNPACK 2cd86bModel: FP16MobileNetV3Smalla2004006008001000SE +/- 20.00, N = 38811. (CXX) g++ options: -O3 -lrt -lm

XNNPACK

Model: QU8MobileNetV2

OpenBenchmarking.orgus, Fewer Is BetterXNNPACK 2cd86bModel: QU8MobileNetV2a2004006008001000SE +/- 6.69, N = 39451. (CXX) g++ options: -O3 -lrt -lm

XNNPACK

Model: QU8MobileNetV3Large

OpenBenchmarking.orgus, Fewer Is BetterXNNPACK 2cd86bModel: QU8MobileNetV3Largea30060090012001500SE +/- 8.97, N = 314841. (CXX) g++ options: -O3 -lrt -lm

XNNPACK

Model: QU8MobileNetV3Small

OpenBenchmarking.orgus, Fewer Is BetterXNNPACK 2cd86bModel: QU8MobileNetV3Smalla2004006008001000SE +/- 9.82, N = 310831. (CXX) g++ options: -O3 -lrt -lm

Phoronix Test Suite v10.8.5