onednn y 3990X AMD Ryzen Threadripper 3990X 64-Core testing with a Gigabyte TRX40 AORUS PRO WIFI (F4p BIOS) and AMD Radeon RX 5700 8GB on Ubuntu 22.10 via the Phoronix Test Suite.
HTML result view exported from: https://openbenchmarking.org/result/2209285-PTS-ONEDNNY340&export=pdf&grr&rdt&rro .
onednn y 3990X Processor Motherboard Chipset Memory Disk Graphics Audio Monitor Network OS Kernel Desktop Display Server OpenGL Vulkan Compiler File-System Screen Resolution A B C AMD Ryzen Threadripper 3990X 64-Core @ 2.90GHz (64 Cores / 128 Threads) Gigabyte TRX40 AORUS PRO WIFI (F4p BIOS) AMD Starship/Matisse 128GB Samsung SSD 970 EVO Plus 500GB AMD Radeon RX 5700 8GB (1750/875MHz) AMD Navi 10 HDMI Audio DELL P2415Q Intel I211 + Intel Wi-Fi 6 AX200 Ubuntu 22.10 6.0.0-060000rc7daily20220927-generic (x86_64) GNOME Shell X Server 1.21.1.3 + Wayland 4.6 Mesa 22.1.7 (LLVM 14.0.6 DRM 3.48) 1.3.211 GCC 12.2.0 ext4 3840x2160 OpenBenchmarking.org Kernel Details - Transparent Huge Pages: madvise Compiler Details - --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --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 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-12-Wbc0TK/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-Wbc0TK/gcc-12-12.2.0/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-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v Processor Details - Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0x8301055 Security Details - itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Mitigation of untrained return thunk; SMT enabled with STIBP protection + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
onednn y 3990X y-cruncher: 10B onednn: Recurrent Neural Network Training - u8s8f32 - CPU onednn: Recurrent Neural Network Training - f32 - CPU onednn: Recurrent Neural Network Training - bf16bf16bf16 - CPU onednn: Recurrent Neural Network Inference - bf16bf16bf16 - CPU onednn: Recurrent Neural Network Inference - u8s8f32 - CPU onednn: Recurrent Neural Network Inference - f32 - CPU y-cruncher: 1B onednn: Deconvolution Batch shapes_1d - f32 - CPU onednn: Deconvolution Batch shapes_1d - u8s8f32 - CPU onednn: IP Shapes 1D - f32 - CPU onednn: IP Shapes 1D - u8s8f32 - CPU y-cruncher: 500M onednn: Matrix Multiply Batch Shapes Transformer - f32 - CPU onednn: Matrix Multiply Batch Shapes Transformer - u8s8f32 - CPU onednn: IP Shapes 3D - f32 - CPU onednn: IP Shapes 3D - u8s8f32 - CPU onednn: Convolution Batch Shapes Auto - f32 - CPU onednn: Convolution Batch Shapes Auto - u8s8f32 - CPU onednn: Deconvolution Batch shapes_3d - f32 - CPU onednn: Deconvolution Batch shapes_3d - u8s8f32 - CPU A B C 277.637 5170.56 5099.39 5084.99 1311.31 1261.59 1290.42 22.616 8.84997 1.78326 2.46105 3.32556 10.999 6.60041 14.3554 5.34789 1.11141 0.963938 6.40545 2.08438 0.972299 277.5 5089.14 5075.05 5055.59 1269.62 1242.98 1252.43 22.53 9.04081 1.87072 1.61291 2.36497 10.972 6.25739 14.3409 5.31681 1.10316 0.850528 6.41234 2.0978 0.996593 277.874 5096.85 5065.26 5074.13 1262.25 1284.38 1329.45 22.589 8.42061 1.92539 1.64163 3.58297 11.027 10.274 13.4715 5.35675 1.10834 0.948633 6.43985 2.08861 0.98403 OpenBenchmarking.org
Y-Cruncher Pi Digits To Calculate: 10B OpenBenchmarking.org Seconds, Fewer Is Better Y-Cruncher 0.7.10.9513 Pi Digits To Calculate: 10B C B A 60 120 180 240 300 277.87 277.50 277.64
oneDNN Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU C B A 1100 2200 3300 4400 5500 5096.85 5089.14 5170.56 MIN: 5046.62 MIN: 5032.11 MIN: 5113.2 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
oneDNN Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU C B A 1100 2200 3300 4400 5500 5065.26 5075.05 5099.39 MIN: 5018.58 MIN: 5024.01 MIN: 5042.36 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
oneDNN Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU C B A 1100 2200 3300 4400 5500 5074.13 5055.59 5084.99 MIN: 5026.36 MIN: 5000.27 MIN: 5029.13 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
oneDNN Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU C B A 300 600 900 1200 1500 1262.25 1269.62 1311.31 MIN: 1236.18 MIN: 1241.95 MIN: 1280.37 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
oneDNN Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU C B A 300 600 900 1200 1500 1284.38 1242.98 1261.59 MIN: 1250.24 MIN: 1218.38 MIN: 1232.39 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
oneDNN Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU C B A 300 600 900 1200 1500 1329.45 1252.43 1290.42 MIN: 1292.26 MIN: 1220.99 MIN: 1257.52 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
Y-Cruncher Pi Digits To Calculate: 1B OpenBenchmarking.org Seconds, Fewer Is Better Y-Cruncher 0.7.10.9513 Pi Digits To Calculate: 1B C B A 5 10 15 20 25 22.59 22.53 22.62
oneDNN Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPU C B A 3 6 9 12 15 8.42061 9.04081 8.84997 MIN: 7.12 MIN: 7.43 MIN: 6.87 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
oneDNN Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU C B A 0.4332 0.8664 1.2996 1.7328 2.166 1.92539 1.87072 1.78326 MIN: 1.58 MIN: 1.64 MIN: 1.54 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
oneDNN Harness: IP Shapes 1D - Data Type: f32 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: IP Shapes 1D - Data Type: f32 - Engine: CPU C B A 0.5537 1.1074 1.6611 2.2148 2.7685 1.64163 1.61291 2.46105 MIN: 1.44 MIN: 1.41 MIN: 2.06 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
oneDNN Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU C B A 0.8062 1.6124 2.4186 3.2248 4.031 3.58297 2.36497 3.32556 MIN: 2.61 MIN: 1.88 MIN: 2.29 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
Y-Cruncher Pi Digits To Calculate: 500M OpenBenchmarking.org Seconds, Fewer Is Better Y-Cruncher 0.7.10.9513 Pi Digits To Calculate: 500M C B A 3 6 9 12 15 11.03 10.97 11.00
oneDNN Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU C B A 3 6 9 12 15 10.27400 6.25739 6.60041 MIN: 9.98 MIN: 6.01 MIN: 6.35 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
oneDNN Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU C B A 4 8 12 16 20 13.47 14.34 14.36 MIN: 12.66 MIN: 13.83 MIN: 13.65 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
oneDNN Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU C B A 1.2053 2.4106 3.6159 4.8212 6.0265 5.35675 5.31681 5.34789 MIN: 5.16 MIN: 5.12 MIN: 5.14 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
oneDNN Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU C B A 0.2501 0.5002 0.7503 1.0004 1.2505 1.10834 1.10316 1.11141 MIN: 1.02 MIN: 1.01 MIN: 1.02 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
oneDNN Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU C B A 0.2169 0.4338 0.6507 0.8676 1.0845 0.948633 0.850528 0.963938 MIN: 0.89 MIN: 0.81 MIN: 0.89 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
oneDNN Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU C B A 2 4 6 8 10 6.43985 6.41234 6.40545 MIN: 6.34 MIN: 6.31 MIN: 6.31 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
oneDNN Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPU C B A 0.472 0.944 1.416 1.888 2.36 2.08861 2.09780 2.08438 MIN: 2.04 MIN: 2.04 MIN: 2.03 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
oneDNN Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU C B A 0.2242 0.4484 0.6726 0.8968 1.121 0.984030 0.996593 0.972299 MIN: 0.93 MIN: 0.93 MIN: 0.92 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
Phoronix Test Suite v10.8.5