onednn onnx alderlake

Intel Core i9-12900K testing with a ASUS ROG STRIX Z690-E GAMING WIFI (1003 BIOS) and Gigabyte AMD Radeon RX 6800 XT 16GB on Ubuntu 21.10 via the Phoronix Test Suite.

HTML result view exported from: https://openbenchmarking.org/result/2203319-NE-ONEDNNONN20&gru&sor.

onednn onnx alderlakeProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerOpenGLVulkanCompilerFile-SystemScreen ResolutionABCDIntel Core i9-12900K @ 5.20GHz (16 Cores / 24 Threads)ASUS ROG STRIX Z690-E GAMING WIFI (1003 BIOS)Intel Device 7aa732GB1000GB Western Digital WDS100T1X0E-00AFY0 + 2000GBGigabyte AMD Radeon RX 6800 XT 16GB (2575/1000MHz)Intel Device 7ad0ASUS VP28UIntel I225-V + Intel Wi-Fi 6 AX210/AX211/AX411Ubuntu 21.105.17.0-phx (x86_64)GNOME Shell 40.5X Server 1.20.13 + Wayland4.6 Mesa 22.1.0-devel (git-ae710f3 2022-03-26 impish-oibaf-ppa) (LLVM 13.0.1 DRM 3.46)1.3.207GCC 11.2.0ext43840x2160OpenBenchmarking.orgKernel Details- Transparent Huge Pages: madviseCompiler Details- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-ZPT0kp/gcc-11-11.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-ZPT0kp/gcc-11-11.2.0/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 Details- Scaling Governor: intel_pstate powersave (EPP: balance_performance) - CPU Microcode: 0x18 - Thermald 2.4.6Python Details- Python 3.9.7Security Details- itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: 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 IBRS IBPB: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected

onednn onnx alderlakeonnx: GPT-2 - CPU - Parallelonnx: GPT-2 - CPU - Standardonnx: yolov4 - CPU - Parallelonnx: yolov4 - CPU - Standardonnx: bertsquad-12 - CPU - Parallelonnx: bertsquad-12 - CPU - Standardonnx: fcn-resnet101-11 - CPU - Parallelonnx: fcn-resnet101-11 - CPU - Standardonnx: ArcFace ResNet-100 - CPU - Parallelonnx: ArcFace ResNet-100 - CPU - Standardonnx: super-resolution-10 - CPU - Parallelonnx: super-resolution-10 - CPU - Standardonednn: IP Shapes 1D - f32 - CPUonednn: IP Shapes 3D - f32 - CPUonednn: IP Shapes 1D - u8s8f32 - CPUonednn: IP Shapes 3D - u8s8f32 - CPUonednn: Convolution Batch Shapes Auto - f32 - CPUonednn: Deconvolution Batch shapes_1d - f32 - CPUonednn: Deconvolution Batch shapes_3d - f32 - CPUonednn: Convolution Batch Shapes Auto - u8s8f32 - CPUonednn: Deconvolution Batch shapes_1d - u8s8f32 - CPUonednn: Deconvolution Batch shapes_3d - u8s8f32 - CPUonednn: Recurrent Neural Network Training - f32 - CPUonednn: Recurrent Neural Network Inference - f32 - CPUonednn: Recurrent Neural Network Training - u8s8f32 - CPUonednn: Recurrent Neural Network Inference - u8s8f32 - CPUonednn: Matrix Multiply Batch Shapes Transformer - f32 - CPUonednn: Recurrent Neural Network Training - bf16bf16bf16 - CPUonednn: Recurrent Neural Network Inference - bf16bf16bf16 - CPUonednn: Matrix Multiply Batch Shapes Transformer - u8s8f32 - CPUonednn: Matrix Multiply Batch Shapes Transformer - bf16bf16bf16 - CPUABCD798911035628670925931111963631908451853322.637243.912131.082790.8820825.908658.276805.245366.054361.353762.217372883.121614.542881.961617.551.245672884.941615.081.033654811011077632664937916111953641913450453492.651353.818871.055780.8516275.903208.384115.246016.046151.353322.220082875.931611.802885.961612.801.255102881.541630.530.826339804611079631666933918111953621896444453282.631803.884871.057540.8817995.905788.214155.246486.056621.340612.220752881.761612.972882.621611.271.284322879.531618.780.842998809511123633669964926111953641893450653842.632873.887171.059230.8850575.909308.050845.248266.047981.342962.219302879.921614.782880.351614.271.310812882.221616.111.011141OpenBenchmarking.org

ONNX Runtime

Model: GPT-2 - Device: CPU - Executor: Parallel

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: GPT-2 - Device: CPU - Executor: ParallelBDCA2K4K6K8K10KSE +/- 47.85, N = 3SE +/- 24.57, N = 3SE +/- 40.10, N = 3SE +/- 58.12, N = 381108095804679891. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: GPT-2 - Device: CPU - Executor: Standard

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: GPT-2 - Device: CPU - Executor: StandardDCBA2K4K6K8K10KSE +/- 41.07, N = 3SE +/- 13.33, N = 3SE +/- 11.10, N = 3SE +/- 13.90, N = 3111231107911077110351. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: yolov4 - Device: CPU - Executor: Parallel

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: yolov4 - Device: CPU - Executor: ParallelDBCA140280420560700SE +/- 1.17, N = 3SE +/- 2.03, N = 3SE +/- 2.05, N = 3SE +/- 2.75, N = 36336326316281. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: yolov4 - Device: CPU - Executor: Standard

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: yolov4 - Device: CPU - Executor: StandardADCB140280420560700SE +/- 2.33, N = 3SE +/- 3.53, N = 3SE +/- 2.74, N = 3SE +/- 4.18, N = 36706696666641. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

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

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: bertsquad-12 - Device: CPU - Executor: ParallelDBCA2004006008001000SE +/- 6.62, N = 3SE +/- 10.12, N = 5SE +/- 12.55, N = 3SE +/- 8.08, N = 39649379339251. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

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

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: bertsquad-12 - Device: CPU - Executor: StandardADCB2004006008001000SE +/- 5.01, N = 3SE +/- 2.60, N = 3SE +/- 9.18, N = 3SE +/- 3.92, N = 39319269189161. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

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

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: fcn-resnet101-11 - Device: CPU - Executor: ParallelDCBA20406080100SE +/- 0.17, N = 3SE +/- 0.17, N = 3SE +/- 0.29, N = 3SE +/- 0.60, N = 31111111111111. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

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

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: fcn-resnet101-11 - Device: CPU - Executor: StandardADCB20406080100SE +/- 0.33, N = 3SE +/- 1.00, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3969595951. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: ArcFace ResNet-100 - Device: CPU - Executor: Parallel

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: ArcFace ResNet-100 - Device: CPU - Executor: ParallelDBAC80160240320400SE +/- 0.29, N = 3SE +/- 0.50, N = 3SE +/- 0.17, N = 3SE +/- 0.17, N = 33643643633621. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

Model: ArcFace ResNet-100 - Device: CPU - Executor: Standard

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: ArcFace ResNet-100 - Device: CPU - Executor: StandardBACD400800120016002000SE +/- 2.93, N = 3SE +/- 5.36, N = 3SE +/- 11.79, N = 3SE +/- 4.07, N = 319131908189618931. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

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

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: super-resolution-10 - Device: CPU - Executor: ParallelADBC10002000300040005000SE +/- 53.96, N = 3SE +/- 57.47, N = 3SE +/- 63.24, N = 3SE +/- 28.86, N = 345184506450444441. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt

ONNX Runtime

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

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: super-resolution-10 - Device: CPU - Executor: StandardDBAC12002400360048006000SE +/- 52.89, N = 3SE +/- 31.64, N = 3SE +/- 20.95, N = 3SE +/- 28.50, N = 353845349533253281. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt

oneDNN

Harness: IP Shapes 1D - Data Type: f32 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: IP Shapes 1D - Data Type: f32 - Engine: CPUCDAB0.59661.19321.78982.38642.983SE +/- 0.00440, N = 3SE +/- 0.00281, N = 3SE +/- 0.00542, N = 3SE +/- 0.01495, N = 32.631802.632872.637242.65135MIN: 2.49MIN: 2.51MIN: 2.5MIN: 2.511. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread

oneDNN

Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: IP Shapes 3D - Data Type: f32 - Engine: CPUBCDA0.88021.76042.64063.52084.401SE +/- 0.00339, N = 3SE +/- 0.00218, N = 3SE +/- 0.00582, N = 3SE +/- 0.00073, N = 33.818873.884873.887173.91213MIN: 3.78MIN: 3.86MIN: 3.85MIN: 3.891. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread

oneDNN

Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPUBCDA0.24360.48720.73080.97441.218SE +/- 0.00867, N = 3SE +/- 0.00884, N = 3SE +/- 0.00217, N = 3SE +/- 0.01605, N = 151.055781.057541.059231.08279MIN: 1.01MIN: 1.01MIN: 1.02MIN: 1.011. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread

oneDNN

Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPUBCAD0.19910.39820.59730.79640.9955SE +/- 0.003743, N = 3SE +/- 0.002403, N = 3SE +/- 0.000523, N = 3SE +/- 0.006655, N = 30.8516270.8817990.8820820.885057MIN: 0.83MIN: 0.86MIN: 0.87MIN: 0.861. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread

oneDNN

Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPUBCAD1.32962.65923.98885.31846.648SE +/- 0.00395, N = 3SE +/- 0.00383, N = 3SE +/- 0.00465, N = 3SE +/- 0.00197, N = 35.903205.905785.908655.90930MIN: 5.83MIN: 5.82MIN: 5.79MIN: 5.751. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread

oneDNN

Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPUDCAB246810SE +/- 0.05693, N = 3SE +/- 0.09831, N = 15SE +/- 0.10247, N = 15SE +/- 0.11946, N = 148.050848.214158.276808.38411MIN: 4.24MIN: 4.09MIN: 4.02MIN: 3.991. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread

oneDNN

Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPUABCD1.18092.36183.54274.72365.9045SE +/- 0.00223, N = 3SE +/- 0.00260, N = 3SE +/- 0.00169, N = 3SE +/- 0.00317, N = 35.245365.246015.246485.24826MIN: 5.18MIN: 5.17MIN: 5.19MIN: 5.171. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread

oneDNN

Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPUBDAC246810SE +/- 0.00195, N = 3SE +/- 0.00248, N = 3SE +/- 0.00353, N = 3SE +/- 0.00242, N = 36.046156.047986.054366.05662MIN: 5.97MIN: 5.95MIN: 5.99MIN: 5.91. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread

oneDNN

Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPUCDBA0.30460.60920.91381.21841.523SE +/- 0.00457, N = 3SE +/- 0.00783, N = 3SE +/- 0.01445, N = 3SE +/- 0.01690, N = 31.340611.342961.353321.35376MIN: 1.28MIN: 1.28MIN: 1.28MIN: 1.281. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread

oneDNN

Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPUADBC0.49970.99941.49911.99882.4985SE +/- 0.00202, N = 3SE +/- 0.00390, N = 3SE +/- 0.00095, N = 3SE +/- 0.00145, N = 32.217372.219302.220082.22075MIN: 2.2MIN: 2.19MIN: 2.19MIN: 2.21. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread

oneDNN

Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPUBDCA6001200180024003000SE +/- 3.97, N = 3SE +/- 3.46, N = 3SE +/- 2.26, N = 3SE +/- 3.43, N = 32875.932879.922881.762883.12MIN: 2865.18MIN: 2866.64MIN: 2866.35MIN: 2869.721. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread

oneDNN

Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPUBCAD30060090012001500SE +/- 2.11, N = 3SE +/- 0.75, N = 3SE +/- 2.06, N = 3SE +/- 1.49, N = 31611.801612.971614.541614.78MIN: 1603.03MIN: 1606.37MIN: 1608.19MIN: 1607.961. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread

oneDNN

Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPUDACB6001200180024003000SE +/- 2.59, N = 3SE +/- 6.09, N = 3SE +/- 1.28, N = 3SE +/- 2.20, N = 32880.352881.962882.622885.96MIN: 2869.74MIN: 2865.7MIN: 2872.99MIN: 2873.231. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread

oneDNN

Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPUCBDA30060090012001500SE +/- 1.60, N = 3SE +/- 1.77, N = 3SE +/- 0.67, N = 3SE +/- 5.04, N = 31611.271612.801614.271617.55MIN: 1602.83MIN: 1604.84MIN: 1608.32MIN: 1606.771. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread

oneDNN

Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPUABCD0.29490.58980.88471.17961.4745SE +/- 0.01358, N = 3SE +/- 0.01551, N = 3SE +/- 0.01243, N = 3SE +/- 0.02113, N = 151.245671.255101.284321.31081MIN: 1.17MIN: 1.17MIN: 1.19MIN: 1.171. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread

oneDNN

Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPUCBDA6001200180024003000SE +/- 2.50, N = 3SE +/- 4.30, N = 3SE +/- 0.97, N = 3SE +/- 4.56, N = 32879.532881.542882.222884.94MIN: 2866.7MIN: 2866.57MIN: 2871.87MIN: 2874.531. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread

oneDNN

Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPUADCB400800120016002000SE +/- 3.13, N = 3SE +/- 2.17, N = 3SE +/- 3.52, N = 3SE +/- 22.13, N = 31615.081616.111618.781630.53MIN: 1605.12MIN: 1608.7MIN: 1609.4MIN: 1601.771. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread

oneDNN

Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPUBCDA0.23260.46520.69780.93041.163SE +/- 0.001870, N = 3SE +/- 0.006474, N = 3SE +/- 0.091776, N = 12SE +/- 0.086658, N = 120.8263390.8429981.0111411.033654MIN: 0.71MIN: 0.72MIN: 0.7MIN: 0.711. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread


Phoronix Test Suite v10.8.4