machine learning threadripper AMD Ryzen Threadripper 3970X 32-Core testing with a ASUS ROG ZENITH II EXTREME (0702 BIOS) and AMD Radeon VII 16GB on Ubuntu 19.10 via the Phoronix Test Suite.
HTML result view exported from: https://openbenchmarking.org/result/2002242-PTS-MACHINEL24&grt .
machine learning threadripper Processor Motherboard Chipset Memory Disk Graphics Audio Monitor Network OS Kernel Desktop Display Server Display Driver OpenGL Compiler File-System Screen Resolution AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core AMD Ryzen Threadripper 3970X 32-Core @ 3.70GHz (32 Cores / 64 Threads) ASUS ROG ZENITH II EXTREME (0702 BIOS) AMD Starship/Matisse 64GB 1000GB Force MP600 AMD Radeon VII 16GB (1802/1001MHz) AMD Vega 20 HDMI Audio ASUS MG28U Aquantia AQC107 NBase-T/IEEE + Intel I211 + Intel Device 2723 Ubuntu 19.10 5.3.0-40-generic (x86_64) GNOME Shell 3.34.1 X Server 1.20.5 amdgpu 19.0.1 4.5 Mesa 19.2.8 (LLVM 9.0.0) GCC 9.2.1 20191008 ext4 3840x2160 OpenBenchmarking.org - --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --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++,gm2 --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-offload-targets=nvptx-none,hsa --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 - Scaling Governor: acpi-cpufreq ondemand - CPU Microcode: 0x8301025 - GLAMOR - Python 2.7.17 + Python 3.7.5 - itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional STIBP: conditional RSB filling + tsx_async_abort: Not affected
machine learning threadripper lczero: BLAS mkl-dnn: IP Batch 1D - f32 mkl-dnn: IP Batch All - f32 mkl-dnn: IP Batch 1D - u8s8f32 mkl-dnn: IP Batch All - u8s8f32 mkl-dnn: Convolution Batch conv_3d - f32 mkl-dnn: Convolution Batch conv_all - f32 mkl-dnn: Convolution Batch conv_3d - u8s8f32 mkl-dnn: Deconvolution Batch deconv_1d - f32 mkl-dnn: Deconvolution Batch deconv_3d - f32 mkl-dnn: Convolution Batch conv_alexnet - f32 mkl-dnn: Convolution Batch conv_all - u8s8f32 mkl-dnn: Deconvolution Batch deconv_all - f32 mkl-dnn: Deconvolution Batch deconv_1d - u8s8f32 mkl-dnn: Deconvolution Batch deconv_3d - u8s8f32 mkl-dnn: Recurrent Neural Network Training - f32 mkl-dnn: Convolution Batch conv_alexnet - u8s8f32 mkl-dnn: Convolution Batch conv_googlenet_v3 - f32 mkl-dnn: Convolution Batch conv_googlenet_v3 - u8s8f32 mlpack: scikit_ica mlpack: scikit_qda mlpack: scikit_svm mlpack: scikit_linearridgeregression numpy: rbenchmark: scikit-learn: AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 2193.94 1.67370 11.0888 20.9272 130.008 4.86201 783.253 3771.87 1.84596 2.67155 99.2820 12473.0 1272.24 1391.81 2331.37 194.659 1380.14 42.3428 692.154 45.25 40.82 11.48 1.44 418.58 0.2837 7.904 OpenBenchmarking.org
LeelaChessZero Backend: BLAS OpenBenchmarking.org Nodes Per Second, More Is Better LeelaChessZero 0.23.2 Backend: BLAS AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 500 1000 1500 2000 2500 SE +/- 28.97, N = 15 2193.94 1. (CXX) g++ options: -lpthread
MKL-DNN DNNL Harness: IP Batch 1D - Data Type: f32 OpenBenchmarking.org ms, Fewer Is Better MKL-DNN DNNL 1.1 Harness: IP Batch 1D - Data Type: f32 AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 0.3766 0.7532 1.1298 1.5064 1.883 SE +/- 0.00773, N = 3 1.67370 MIN: 1.59 1. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
MKL-DNN DNNL Harness: IP Batch All - Data Type: f32 OpenBenchmarking.org ms, Fewer Is Better MKL-DNN DNNL 1.1 Harness: IP Batch All - Data Type: f32 AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 3 6 9 12 15 SE +/- 0.05, N = 3 11.09 MIN: 10.5 1. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
MKL-DNN DNNL Harness: IP Batch 1D - Data Type: u8s8f32 OpenBenchmarking.org ms, Fewer Is Better MKL-DNN DNNL 1.1 Harness: IP Batch 1D - Data Type: u8s8f32 AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 5 10 15 20 25 SE +/- 0.03, N = 3 20.93 MIN: 20.19 1. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
MKL-DNN DNNL Harness: IP Batch All - Data Type: u8s8f32 OpenBenchmarking.org ms, Fewer Is Better MKL-DNN DNNL 1.1 Harness: IP Batch All - Data Type: u8s8f32 AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 30 60 90 120 150 SE +/- 0.18, N = 3 130.01 MIN: 126.3 1. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
MKL-DNN DNNL Harness: Convolution Batch conv_3d - Data Type: f32 OpenBenchmarking.org ms, Fewer Is Better MKL-DNN DNNL 1.1 Harness: Convolution Batch conv_3d - Data Type: f32 AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 1.094 2.188 3.282 4.376 5.47 SE +/- 0.00956, N = 3 4.86201 MIN: 4.69 1. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
MKL-DNN DNNL Harness: Convolution Batch conv_all - Data Type: f32 OpenBenchmarking.org ms, Fewer Is Better MKL-DNN DNNL 1.1 Harness: Convolution Batch conv_all - Data Type: f32 AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 200 400 600 800 1000 SE +/- 4.73, N = 3 783.25 MIN: 771.68 1. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
MKL-DNN DNNL Harness: Convolution Batch conv_3d - Data Type: u8s8f32 OpenBenchmarking.org ms, Fewer Is Better MKL-DNN DNNL 1.1 Harness: Convolution Batch conv_3d - Data Type: u8s8f32 AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 800 1600 2400 3200 4000 SE +/- 3.72, N = 3 3771.87 MIN: 3759.53 1. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
MKL-DNN DNNL Harness: Deconvolution Batch deconv_1d - Data Type: f32 OpenBenchmarking.org ms, Fewer Is Better MKL-DNN DNNL 1.1 Harness: Deconvolution Batch deconv_1d - Data Type: f32 AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 0.4153 0.8306 1.2459 1.6612 2.0765 SE +/- 0.00210, N = 3 1.84596 MIN: 1.8 1. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
MKL-DNN DNNL Harness: Deconvolution Batch deconv_3d - Data Type: f32 OpenBenchmarking.org ms, Fewer Is Better MKL-DNN DNNL 1.1 Harness: Deconvolution Batch deconv_3d - Data Type: f32 AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 0.6011 1.2022 1.8033 2.4044 3.0055 SE +/- 0.02056, N = 3 2.67155 MIN: 2.58 1. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
MKL-DNN DNNL Harness: Convolution Batch conv_alexnet - Data Type: f32 OpenBenchmarking.org ms, Fewer Is Better MKL-DNN DNNL 1.1 Harness: Convolution Batch conv_alexnet - Data Type: f32 AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 20 40 60 80 100 SE +/- 0.60, N = 3 99.28 MIN: 97.33 1. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
MKL-DNN DNNL Harness: Convolution Batch conv_all - Data Type: u8s8f32 OpenBenchmarking.org ms, Fewer Is Better MKL-DNN DNNL 1.1 Harness: Convolution Batch conv_all - Data Type: u8s8f32 AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 3K 6K 9K 12K 15K SE +/- 7.92, N = 3 12473.0 MIN: 12331 1. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
MKL-DNN DNNL Harness: Deconvolution Batch deconv_all - Data Type: f32 OpenBenchmarking.org ms, Fewer Is Better MKL-DNN DNNL 1.1 Harness: Deconvolution Batch deconv_all - Data Type: f32 AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 300 600 900 1200 1500 SE +/- 3.81, N = 3 1272.24 MIN: 1250.51 1. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
MKL-DNN DNNL Harness: Deconvolution Batch deconv_1d - Data Type: u8s8f32 OpenBenchmarking.org ms, Fewer Is Better MKL-DNN DNNL 1.1 Harness: Deconvolution Batch deconv_1d - Data Type: u8s8f32 AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 300 600 900 1200 1500 SE +/- 0.99, N = 3 1391.81 MIN: 1385.92 1. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
MKL-DNN DNNL Harness: Deconvolution Batch deconv_3d - Data Type: u8s8f32 OpenBenchmarking.org ms, Fewer Is Better MKL-DNN DNNL 1.1 Harness: Deconvolution Batch deconv_3d - Data Type: u8s8f32 AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 500 1000 1500 2000 2500 SE +/- 14.76, N = 3 2331.37 MIN: 2300.36 1. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
MKL-DNN DNNL Harness: Recurrent Neural Network Training - Data Type: f32 OpenBenchmarking.org ms, Fewer Is Better MKL-DNN DNNL 1.1 Harness: Recurrent Neural Network Training - Data Type: f32 AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 40 80 120 160 200 SE +/- 1.39, N = 14 194.66 MIN: 190.33 1. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
MKL-DNN DNNL Harness: Convolution Batch conv_alexnet - Data Type: u8s8f32 OpenBenchmarking.org ms, Fewer Is Better MKL-DNN DNNL 1.1 Harness: Convolution Batch conv_alexnet - Data Type: u8s8f32 AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 300 600 900 1200 1500 SE +/- 4.23, N = 3 1380.14 MIN: 1356.84 1. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
MKL-DNN DNNL Harness: Convolution Batch conv_googlenet_v3 - Data Type: f32 OpenBenchmarking.org ms, Fewer Is Better MKL-DNN DNNL 1.1 Harness: Convolution Batch conv_googlenet_v3 - Data Type: f32 AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 10 20 30 40 50 SE +/- 0.28, N = 3 42.34 MIN: 41.48 1. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
MKL-DNN DNNL Harness: Convolution Batch conv_googlenet_v3 - Data Type: u8s8f32 OpenBenchmarking.org ms, Fewer Is Better MKL-DNN DNNL 1.1 Harness: Convolution Batch conv_googlenet_v3 - Data Type: u8s8f32 AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 150 300 450 600 750 SE +/- 2.69, N = 2 692.15 MIN: 675.46 1. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
Mlpack Benchmark Benchmark: scikit_ica OpenBenchmarking.org Seconds, Fewer Is Better Mlpack Benchmark Benchmark: scikit_ica AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 10 20 30 40 50 SE +/- 0.06, N = 3 45.25
Mlpack Benchmark Benchmark: scikit_qda OpenBenchmarking.org Seconds, Fewer Is Better Mlpack Benchmark Benchmark: scikit_qda AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 9 18 27 36 45 SE +/- 0.42, N = 3 40.82
Mlpack Benchmark Benchmark: scikit_svm OpenBenchmarking.org Seconds, Fewer Is Better Mlpack Benchmark Benchmark: scikit_svm AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 3 6 9 12 15 SE +/- 0.03, N = 3 11.48
Mlpack Benchmark Benchmark: scikit_linearridgeregression OpenBenchmarking.org Seconds, Fewer Is Better Mlpack Benchmark Benchmark: scikit_linearridgeregression AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 0.324 0.648 0.972 1.296 1.62 SE +/- 0.02, N = 3 1.44
Numpy Benchmark OpenBenchmarking.org Score, More Is Better Numpy Benchmark AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 90 180 270 360 450 SE +/- 2.37, N = 3 418.58
R Benchmark OpenBenchmarking.org Seconds, Fewer Is Better R Benchmark AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 0.0638 0.1276 0.1914 0.2552 0.319 SE +/- 0.0011, N = 3 0.2837 1. R scripting front-end version 3.6.1 (2019-07-05)
Scikit-Learn OpenBenchmarking.org Seconds, Fewer Is Better Scikit-Learn 0.22.1 AMD Radeon VII - AMD Ryzen Threadripper 3970X 32-Core 2 4 6 8 10 SE +/- 0.015, N = 3 7.904
Phoronix Test Suite v10.8.4