Xeon E onednn

Intel Xeon E-2288G testing with a Compulab SBC-ATCFL v1.2 (ATOP3.PRD.0.29.2 BIOS) and NVIDIA Quadro RTX 4000 8GB on Ubuntu 20.10 via the Phoronix Test Suite.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2103135-FI-XEONEONED91
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

Statistics

Show Overall Harmonic Mean(s)
Show Overall Geometric Mean
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

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
View Logs
Performance Per
Dollar
Date
Run
  Test
  Duration
1
March 13 2021
  41 Minutes
2
March 13 2021
  35 Minutes
3
March 13 2021
  41 Minutes
Invert Hiding All Results Option
  39 Minutes

Only show results where is faster than
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):

1

Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --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-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-gcn/usr,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
Processor Notes: Scaling Governor: intel_pstate powersave - CPU Microcode: 0xde
Security Notes: itlb_multihit: KVM: Mitigation of VMX disabled + 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 Enhanced IBRS IBPB: conditional RSB filling + srbds: Mitigation of TSX disabled + tsx_async_abort: Mitigation of TSX disabled

2

3

Processor: Intel Xeon E-2288G @ 5.00GHz (8 Cores / 16 Threads), Motherboard: Compulab SBC-ATCFL v1.2 (ATOP3.PRD.0.29.2 BIOS), Chipset: Intel Cannon Lake PCH, Memory: 2 x 32 GB DDR4-2667MT/s Samsung M378A4G43MB1-CTD, Disk: Samsung SSD 970 EVO Plus 250GB, Graphics: NVIDIA Quadro RTX 4000 8GB, Audio: Intel Cannon Lake PCH cAVS, Network: Intel I219-LM + Intel I210

OS: Ubuntu 20.10, Kernel: 5.8.0-41-generic (x86_64), Desktop: GNOME Shell 3.38.2, Display Server: X Server 1.20.9, Display Driver: modesetting 1.20.9, OpenCL: OpenCL 1.2 CUDA 11.2.109, Vulkan: 1.2.155, Compiler: GCC 10.2.0, File-System: ext4, Screen Resolution: 1920x1080

Xeon E onednnProcessorMotherboardChipsetMemoryDiskGraphicsAudioNetworkOSKernelDesktopDisplay ServerDisplay DriverOpenCLVulkanCompilerFile-SystemScreen Resolution123Intel Xeon E-2288G @ 5.00GHz (8 Cores / 16 Threads)Compulab SBC-ATCFL v1.2 (ATOP3.PRD.0.29.2 BIOS)Intel Cannon Lake PCH2 x 32 GB DDR4-2667MT/s Samsung M378A4G43MB1-CTDSamsung SSD 970 EVO Plus 250GBNVIDIA Quadro RTX 4000 8GBIntel Cannon Lake PCH cAVSIntel I219-LM + Intel I210Ubuntu 20.105.8.0-41-generic (x86_64)GNOME Shell 3.38.2X Server 1.20.9modesetting 1.20.9OpenCL 1.2 CUDA 11.2.1091.2.155GCC 10.2.0ext41920x1080OpenBenchmarking.orgCompiler Details- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --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-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-gcn/usr,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 Processor Details- Scaling Governor: intel_pstate powersave - CPU Microcode: 0xdeSecurity Details- itlb_multihit: KVM: Mitigation of VMX disabled + 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 Enhanced IBRS IBPB: conditional RSB filling + srbds: Mitigation of TSX disabled + tsx_async_abort: Mitigation of TSX disabled

123Result OverviewPhoronix Test Suite100%101%102%104%105%oneDNNoneDNNoneDNNoneDNNoneDNNoneDNNoneDNNoneDNNoneDNNoneDNNoneDNNoneDNNoneDNNoneDNNoneDNNoneDNNoneDNNoneDNND.B.s - f32 - CPUIP Shapes 1D - u8s8f32 - CPUIP Shapes 1D - f32 - CPUIP Shapes 3D - u8s8f32 - CPUIP Shapes 3D - f32 - CPUD.B.s - u8s8f32 - CPUD.B.s - f32 - CPUC.B.S.A - f32 - CPUC.B.S.A - u8s8f32 - CPUR.N.N.T - bf16bf16bf16 - CPUD.B.s - u8s8f32 - CPUR.N.N.I - f32 - CPUM.M.B.S.T - u8s8f32 - CPUR.N.N.I - bf16bf16bf16 - CPUR.N.N.I - u8s8f32 - CPUM.M.B.S.T - f32 - CPUR.N.N.T - u8s8f32 - CPUR.N.N.T - f32 - CPU

Xeon E onednnonednn: 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 - CPU1234.2772010.89911.874972.4388617.93478.539536.9577417.26392.481263.890094179.632308.214194.522301.834.224024192.732304.723.858444.3948010.93731.962862.4068117.84928.356646.9270917.21052.481463.915464185.772314.854186.182312.664.238614188.732319.553.887064.3470110.80401.936832.4470018.00868.766036.9895517.11212.509343.920784182.362325.804189.692313.224.239554224.762318.113.87967OpenBenchmarking.org

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI initiative. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 1D - Data Type: f32 - Engine: CPU1230.98881.97762.96643.95524.944SE +/- 0.03226, N = 15SE +/- 0.04781, N = 15SE +/- 0.03852, N = 154.277204.394804.34701MIN: 3.59MIN: 3.59MIN: 3.61. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 1D - Data Type: f32 - Engine: CPU123246810Min: 3.94 / Avg: 4.28 / Max: 4.37Min: 3.94 / Avg: 4.39 / Max: 4.59Min: 3.93 / Avg: 4.35 / Max: 4.621. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU1233691215SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.07, N = 310.9010.9410.80MIN: 9.99MIN: 9.98MIN: 9.971. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU1233691215Min: 10.87 / Avg: 10.9 / Max: 10.93Min: 10.88 / Avg: 10.94 / Max: 10.97Min: 10.67 / Avg: 10.8 / Max: 10.911. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU1230.44160.88321.32481.76642.208SE +/- 0.02166, N = 6SE +/- 0.00219, N = 3SE +/- 0.01631, N = 151.874971.962861.93683MIN: 1.55MIN: 1.53MIN: 1.531. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU123246810Min: 1.78 / Avg: 1.87 / Max: 1.94Min: 1.96 / Avg: 1.96 / Max: 1.97Min: 1.77 / Avg: 1.94 / Max: 2.011. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU1230.55061.10121.65182.20242.753SE +/- 0.02338, N = 9SE +/- 0.01956, N = 3SE +/- 0.02297, N = 32.438862.406812.44700MIN: 2.2MIN: 2.2MIN: 2.221. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU123246810Min: 2.41 / Avg: 2.44 / Max: 2.62Min: 2.37 / Avg: 2.41 / Max: 2.44Min: 2.41 / Avg: 2.45 / Max: 2.491. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU12348121620SE +/- 0.12, N = 3SE +/- 0.09, N = 3SE +/- 0.07, N = 317.9317.8518.01MIN: 17.08MIN: 17.12MIN: 17.161. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU123510152025Min: 17.81 / Avg: 17.93 / Max: 18.17Min: 17.72 / Avg: 17.85 / Max: 18.02Min: 17.88 / Avg: 18.01 / Max: 18.131. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPU123246810SE +/- 0.11858, N = 15SE +/- 0.00541, N = 3SE +/- 0.15702, N = 128.539538.356648.76603MIN: 4.45MIN: 5.06MIN: 4.841. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPU1233691215Min: 7.61 / Avg: 8.54 / Max: 9.31Min: 8.35 / Avg: 8.36 / Max: 8.37Min: 8.18 / Avg: 8.77 / Max: 9.461. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPU123246810SE +/- 0.01715, N = 3SE +/- 0.02867, N = 3SE +/- 0.10068, N = 46.957746.927096.98955MIN: 6.58MIN: 6.55MIN: 6.571. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPU1233691215Min: 6.94 / Avg: 6.96 / Max: 6.99Min: 6.88 / Avg: 6.93 / Max: 6.97Min: 6.8 / Avg: 6.99 / Max: 7.241. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU12348121620SE +/- 0.02, N = 3SE +/- 0.06, N = 3SE +/- 0.01, N = 317.2617.2117.11MIN: 16.39MIN: 16.35MIN: 16.371. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU12348121620Min: 17.24 / Avg: 17.26 / Max: 17.3Min: 17.13 / Avg: 17.21 / Max: 17.34Min: 17.09 / Avg: 17.11 / Max: 17.131. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU1230.56461.12921.69382.25842.823SE +/- 0.00606, N = 3SE +/- 0.00925, N = 3SE +/- 0.01499, N = 32.481262.481462.50934MIN: 2.21MIN: 2.22MIN: 2.081. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU123246810Min: 2.47 / Avg: 2.48 / Max: 2.49Min: 2.46 / Avg: 2.48 / Max: 2.49Min: 2.48 / Avg: 2.51 / Max: 2.531. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU1230.88221.76442.64663.52884.411SE +/- 0.02275, N = 3SE +/- 0.02302, N = 3SE +/- 0.00661, N = 33.890093.915463.92078MIN: 3.66MIN: 3.68MIN: 3.691. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU123246810Min: 3.86 / Avg: 3.89 / Max: 3.93Min: 3.87 / Avg: 3.92 / Max: 3.95Min: 3.91 / Avg: 3.92 / Max: 3.931. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU1239001800270036004500SE +/- 19.54, N = 3SE +/- 5.52, N = 3SE +/- 8.62, N = 34179.634185.774182.36MIN: 4110.36MIN: 4137.54MIN: 4131.321. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU1237001400210028003500Min: 4142.89 / Avg: 4179.63 / Max: 4209.55Min: 4175.44 / Avg: 4185.77 / Max: 4194.3Min: 4168.38 / Avg: 4182.36 / Max: 4198.11. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU1235001000150020002500SE +/- 2.57, N = 3SE +/- 13.89, N = 3SE +/- 15.48, N = 32308.212314.852325.80MIN: 2273.4MIN: 2265.89MIN: 2264.931. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU123400800120016002000Min: 2305.05 / Avg: 2308.21 / Max: 2313.3Min: 2293.98 / Avg: 2314.85 / Max: 2341.16Min: 2303.4 / Avg: 2325.8 / Max: 2355.511. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU1239001800270036004500SE +/- 25.41, N = 3SE +/- 3.89, N = 3SE +/- 4.37, N = 34194.524186.184189.69MIN: 4124.44MIN: 4137.73MIN: 4128.171. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU1237001400210028003500Min: 4161.57 / Avg: 4194.52 / Max: 4244.5Min: 4178.62 / Avg: 4186.18 / Max: 4191.56Min: 4183.39 / Avg: 4189.69 / Max: 4198.091. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU1235001000150020002500SE +/- 5.51, N = 3SE +/- 4.70, N = 3SE +/- 3.66, N = 32301.832312.662313.22MIN: 2256.72MIN: 2275.31MIN: 2276.221. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU123400800120016002000Min: 2294.54 / Avg: 2301.83 / Max: 2312.63Min: 2305.83 / Avg: 2312.66 / Max: 2321.66Min: 2308.36 / Avg: 2313.22 / Max: 2320.391. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU1230.95391.90782.86173.81564.7695SE +/- 0.01514, N = 3SE +/- 0.00373, N = 3SE +/- 0.03388, N = 34.224024.238614.23955MIN: 3.84MIN: 3.86MIN: 3.841. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU123246810Min: 4.2 / Avg: 4.22 / Max: 4.25Min: 4.23 / Avg: 4.24 / Max: 4.25Min: 4.2 / Avg: 4.24 / Max: 4.311. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU1239001800270036004500SE +/- 15.97, N = 3SE +/- 7.85, N = 3SE +/- 23.95, N = 34192.734188.734224.76MIN: 4116.8MIN: 4141.92MIN: 4158.661. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU1237001400210028003500Min: 4161.38 / Avg: 4192.73 / Max: 4213.65Min: 4174.2 / Avg: 4188.73 / Max: 4201.16Min: 4193.91 / Avg: 4224.76 / Max: 4271.931. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU1235001000150020002500SE +/- 0.71, N = 3SE +/- 11.13, N = 3SE +/- 4.06, N = 32304.722319.552318.11MIN: 2267.17MIN: 2266.99MIN: 22711. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU123400800120016002000Min: 2303.43 / Avg: 2304.72 / Max: 2305.89Min: 2306.97 / Avg: 2319.55 / Max: 2341.74Min: 2311.4 / Avg: 2318.11 / Max: 2325.411. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU1230.87461.74922.62383.49844.373SE +/- 0.00671, N = 3SE +/- 0.01710, N = 3SE +/- 0.00607, N = 33.858443.887063.87967MIN: 3.24MIN: 3.17MIN: 3.241. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU123246810Min: 3.85 / Avg: 3.86 / Max: 3.87Min: 3.86 / Avg: 3.89 / Max: 3.92Min: 3.87 / Avg: 3.88 / Max: 3.891. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

1

Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --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-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-gcn/usr,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
Processor Notes: Scaling Governor: intel_pstate powersave - CPU Microcode: 0xde
Security Notes: itlb_multihit: KVM: Mitigation of VMX disabled + 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 Enhanced IBRS IBPB: conditional RSB filling + srbds: Mitigation of TSX disabled + tsx_async_abort: Mitigation of TSX disabled

Testing initiated at 13 March 2021 09:55 by user root.

2

Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --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-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-gcn/usr,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
Processor Notes: Scaling Governor: intel_pstate powersave - CPU Microcode: 0xde
Security Notes: itlb_multihit: KVM: Mitigation of VMX disabled + 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 Enhanced IBRS IBPB: conditional RSB filling + srbds: Mitigation of TSX disabled + tsx_async_abort: Mitigation of TSX disabled

Testing initiated at 13 March 2021 10:47 by user root.

3

Processor: Intel Xeon E-2288G @ 5.00GHz (8 Cores / 16 Threads), Motherboard: Compulab SBC-ATCFL v1.2 (ATOP3.PRD.0.29.2 BIOS), Chipset: Intel Cannon Lake PCH, Memory: 2 x 32 GB DDR4-2667MT/s Samsung M378A4G43MB1-CTD, Disk: Samsung SSD 970 EVO Plus 250GB, Graphics: NVIDIA Quadro RTX 4000 8GB, Audio: Intel Cannon Lake PCH cAVS, Network: Intel I219-LM + Intel I210

OS: Ubuntu 20.10, Kernel: 5.8.0-41-generic (x86_64), Desktop: GNOME Shell 3.38.2, Display Server: X Server 1.20.9, Display Driver: modesetting 1.20.9, OpenCL: OpenCL 1.2 CUDA 11.2.109, Vulkan: 1.2.155, Compiler: GCC 10.2.0, File-System: ext4, Screen Resolution: 1920x1080

Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --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-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-gcn/usr,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
Processor Notes: Scaling Governor: intel_pstate powersave - CPU Microcode: 0xde
Security Notes: itlb_multihit: KVM: Mitigation of VMX disabled + 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 Enhanced IBRS IBPB: conditional RSB filling + srbds: Mitigation of TSX disabled + tsx_async_abort: Mitigation of TSX disabled

Testing initiated at 13 March 2021 11:34 by user root.