Threadripper EOY2019 Clang vs. GCC

AMD Ryzen Threadripper 3960X GCC vs. LLVM Clang compiler benchmarks by Michael Larabel for a future article.

HTML result view exported from: https://openbenchmarking.org/result/1912235-PTS-THREADRI74&sor&grs.

Threadripper EOY2019 Clang vs. GCCProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerDisplay DriverOpenGLCompilerFile-SystemScreen ResolutionGCC 10.0.0 20191208LLVM Clang 10.0 20191222GCC 9.2.1LLVM Clang 9.0.0AMD Ryzen Threadripper 3960X 24-Core @ 3.80GHz (24 Cores / 48 Threads)MSI Creator TRX40 (MS-7C59) v1.0 (1.12N1 BIOS)AMD Starship/Matisse32768MB1000GB Sabrent Rocket 4.0 1TBGigabyte AMD Radeon 540/540X/550/550X / RX 540X/550/550X 2GB (1206/1750MHz)AMD Baffin HDMI/DPASUS VP28UAquantia AQC107 NBase-T/IEEE + Intel I211 + Intel Device 2723Ubuntu 19.105.4.0-nvme-hwmon (x86_64)GNOME Shell 3.34.1X Server 1.20.5modesetting 1.20.54.5 Mesa 19.2.1 (LLVM 9.0.0)GCC 10.0.0 20191208ext43840x2160Clang 10.0.0GCC 9.2.1 20191008Clang 9.0.0OpenBenchmarking.orgEnvironment Details- CXXFLAGS="-O3 -march=native" CFLAGS="-O3 -march=native"Compiler Details- GCC 10.0.0 20191208: --disable-multilib --enable-checking=release- LLVM Clang 10.0 20191222: Optimized build; Default target: x86_64-unknown-linux-gnu; Host CPU: znver2- GCC 9.2.1: --disable-multilib --enable-checking=releaseDisk Details- NONE / errors=remount-ro,relatime,rwProcessor Details- Scaling Governor: acpi-cpufreq ondemand - CPU Microcode: 0x8301025Python Details- Python 2.7.17rc1 + Python 3.7.5Security Details- 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

Threadripper EOY2019 Clang vs. GCCjohn-the-ripper: Blowfishttsiod-renderer: Phong Rendering With Soft-Shadow Mappingmt-dgemm: Sustained Floating-Point Ratebuild-php: Time To Compilec-ray: Total Time - 4K, 16 Rays Per Pixelbuild-imagemagick: Time To Compileopenssl: RSA 4096-bit Performancedav1d: Chimera 1080p 10-bitmkl-dnn: Recurrent Neural Network Training - f32parboil: OpenMP MRI Griddingcpp-perf-bench: Rand Numberstungsten: Non-Exponentialqmcpack: fftw: Float + SSE - 2D FFT Size 4096encode-mp3: WAV To MP3cpp-perf-bench: Stepanov Vectoraobench: 2048 x 2048 - Total Timetscp: AI Chess Performancelibgav1: Summer Nature 4Kencode-flac: WAV To FLAClibgav1: Chimera 1080plibgav1: Summer Nature 1080ptungsten: Hairmkl-dnn: IP Batch 1D - f32cpp-perf-bench: Ctypetungsten: Volumetric Causticlibgav1: Chimera 1080p 10-bitcpp-perf-bench: Stepanov Abstractionaskap: tConvolve OpenMP - Degriddingrocksdb: Rand Fillmrbayes: Primate Phylogeny Analysishimeno: Poisson Pressure Solverfftw: Stock - 2D FFT Size 4096aom-av1: AV1 Video Encodingcpp-perf-bench: Function Objectssqlite-speedtest: Timed Time - Size 1,000dav1d: Chimera 1080ptungsten: Water Causticvpxenc: vpxenc VP9 1080p Video Encodestockfish: Total Timerocksdb: Seq Filllczero: Randrodinia: OpenMP CFD Solvermkl-dnn: Convolution Batch conv_alexnet - f32askap: tConvolve OpenMP - Griddingcpp-perf-bench: Math Libraryx265: H.265 1080p Video Encodingminife: Smallpgbench: Buffer Test - Normal Load - Read Onlycompress-zstd: Compressing ubuntu-16.04.3-server-i386.img, Compression Level 19dav1d: Summer Nature 1080pdav1d: Summer Nature 4Kmkl-dnn: Convolution Batch conv_googlenet_v3 - f32rocksdb: Rand Readbyte: Dhrystone 2compress-xz: Compressing ubuntu-16.04.3-server-i386.img, Compression Level 9nginx: Static Web Page Servingx264: H.264 Video Encodingcpp-perf-bench: Atoln-queens: Elapsed Timeaskap: tConvolve MT - Degriddingcrafty: Elapsed Timeaskap: tConvolve MT - Griddingapache: Static Web Page Servingrocksdb: Rand Fill Syncsqlite: 1parboil: OpenMP Stencilrodinia: OpenMP LavaMDrocksdb: Read While Writinggromacs: Water Benchmarkparboil: OpenMP CUTCPpgbench: Buffer Test - Normal Load - Read Writesmallpt: Global Illumination Renderer; 128 Samplesrodinia: OpenMP Streamclusterlczero: BLASGCC 10.0.0 20191208LLVM Clang 10.0 20191222GCC 9.2.1LLVM Clang 9.0.041784961.9748.86318350.24219.13415.0647173.794.47194.07762.9416571061.4803.369271878.1202876.70176.26327.389134599123.898.04251.7077.929.449601.8613132.3113.9669721.9328.2953716.3393089769.7864898.1708218135.20.1513.83857.116612.9418.2157195.638090908210247331058059.185124.2825255.51262.96865.747767.98673644.95828510.052676.37289.2952.469114536896748002215.120.04243035.46199.2258.2784.3783339.3390278351943.5834157.762427714.2357.55170610.08549459162.5151.25741015513.4404234.77519.06929.3289193278.17830.83920.9055189.174.451282.0334.051767.78866.78231.745152824821.177.14546.6970.288.7262529.7953.7248823.1125.86470.7064866.9034270.1514.76459.906625.1119.1580207.68258.40467.11668836.60545310.107685.40291.7619.68743580.60196.1057.84334327.9014.24814868.67884941482935.1238.38597244.93919.15513.5057178.899.72194.76649.0009651069.4833.374191893.6241846.72977.50128.620137308923.748.07351.8477.119.709521.7371032.2204.0790222.0128.5234096.25102333570.3304583.3462128111.60.1414.81656.169587.2419.1975197.377673746210836491059969.151125.0085509.3269.48464.697777.50654500.67635010.040667.33285.2852.282014228776648969411.219.75343426.72197.9357.4734.4293376.2789599371955.9734228.502436814.2527.57639610.10149508902.5161.25771630910.9917853.49219.76132.75986058669.85220.79140364.76330.97213.5317191.673.36145.8501296.9844.070432262.8235597.95167.10831.585148700123.947.19252.4778.578.700241.6731529.0853.6818624.1926.03364.8544983.2817867531.20.1514.69858.332623.0519.3485207.7881336054100236.89.633118.989260.63866.278037.33675821.0958339.817674.39288.6051.328419.68543783.30198.2957.99534144.3914.20031491.37719018.81833.3033OpenBenchmarking.org

John The Ripper

Test: Blowfish

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 1.9.0-jumbo-1Test: BlowfishLLVM Clang 9.0.0GCC 10.0.0 20191208GCC 9.2.1LLVM Clang 10.0 2019122213K26K39K52K65KSE +/- 246.24, N = 3SE +/- 170.74, N = 3SE +/- 184.04, N = 3SE +/- 3.28, N = 36058641784414821932-fopenmp-fopenmp-fopenmp1. (CC) gcc options: -m64 -lssl -lcrypto -lgmp -pthread -lm -lz -ldl -lcrypt

TTSIOD 3D Renderer

Phong Rendering With Soft-Shadow Mapping

OpenBenchmarking.orgFPS, More Is BetterTTSIOD 3D Renderer 2.3bPhong Rendering With Soft-Shadow MappingGCC 10.0.0 20191208GCC 9.2.1LLVM Clang 9.0.02004006008001000SE +/- 1.33, N = 3SE +/- 2.99, N = 3SE +/- 0.09, N = 3961.97935.1269.851. (CXX) g++ options: -O3 -march=native -fomit-frame-pointer -ffast-math -mtune=native -flto -msse -mrecip -mfpmath=sse -msse2 -mssse3 -lSDL -fopenmp -fwhole-program -lstdc++

ACES DGEMM

Sustained Floating-Point Rate

OpenBenchmarking.orgGFLOP/s, More Is BetterACES DGEMM 1.0Sustained Floating-Point RateGCC 10.0.0 20191208GCC 9.2.1LLVM Clang 9.0.0246810SE +/- 0.059593, N = 3SE +/- 0.037368, N = 3SE +/- 0.007958, N = 38.8631838.3859720.7914031. (CC) gcc options: -O3 -march=native -fopenmp

Timed PHP Compilation

Time To Compile

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed PHP Compilation 7.1.9Time To CompileGCC 9.2.1GCC 10.0.0 20191208LLVM Clang 9.0.0LLVM Clang 10.0 2019122220406080100SE +/- 0.08, N = 3SE +/- 0.08, N = 3SE +/- 0.24, N = 3SE +/- 0.12, N = 344.9450.2464.7678.181. (CC) gcc options: -O3 -march=native -pedantic -ldl -lz -lm

C-Ray

Total Time - 4K, 16 Rays Per Pixel

OpenBenchmarking.orgSeconds, Fewer Is BetterC-Ray 1.1Total Time - 4K, 16 Rays Per PixelGCC 10.0.0 20191208GCC 9.2.1LLVM Clang 10.0 20191222LLVM Clang 9.0.0714212835SE +/- 0.05, N = 3SE +/- 0.05, N = 3SE +/- 0.13, N = 3SE +/- 0.11, N = 319.1319.1630.8430.971. (CC) gcc options: -lm -lpthread -O3 -march=native

Timed ImageMagick Compilation

Time To Compile

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed ImageMagick Compilation 6.9.0Time To CompileGCC 9.2.1LLVM Clang 9.0.0GCC 10.0.0 20191208LLVM Clang 10.0 20191222510152025SE +/- 0.08, N = 3SE +/- 0.07, N = 3SE +/- 0.04, N = 3SE +/- 0.08, N = 313.5113.5315.0620.91

OpenSSL

RSA 4096-bit Performance

OpenBenchmarking.orgSigns Per Second, More Is BetterOpenSSL 1.1.1RSA 4096-bit PerformanceLLVM Clang 9.0.0GCC 9.2.1GCC 10.0.0 20191208LLVM Clang 10.0 2019122215003000450060007500SE +/- 29.40, N = 3SE +/- 21.05, N = 3SE +/- 20.75, N = 3SE +/- 21.60, N = 37191.67178.87173.75189.1-Qunused-arguments-Qunused-arguments1. (CC) gcc options: -pthread -m64 -O3 -march=native -lssl -lcrypto -ldl

dav1d

Video Input: Chimera 1080p 10-bit

OpenBenchmarking.orgFPS, More Is Betterdav1d 0.5.0Video Input: Chimera 1080p 10-bitGCC 9.2.1GCC 10.0.0 20191208LLVM Clang 10.0 20191222LLVM Clang 9.0.020406080100SE +/- 0.24, N = 3SE +/- 0.13, N = 3SE +/- 0.12, N = 3SE +/- 0.07, N = 399.7294.4774.4573.36MIN: 60.51 / MAX: 199.48MIN: 56.43 / MAX: 192.99MIN: 45.96 / MAX: 154.32MIN: 45.63 / MAX: 149.131. (CC) gcc options: -O3 -march=native -pthread

MKL-DNN DNNL

Harness: Recurrent Neural Network Training - Data Type: f32

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN DNNL 1.1Harness: Recurrent Neural Network Training - Data Type: f32LLVM Clang 9.0.0GCC 10.0.0 20191208GCC 9.2.14080120160200SE +/- 0.38, N = 3SE +/- 0.39, N = 3SE +/- 0.16, N = 3145.85194.08194.77-fopenmp=libomp - MIN: 143.95-fopenmp - MIN: 192.29-fopenmp - MIN: 192.971. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -pie -lpthread -ldl

Parboil

Test: OpenMP MRI Gridding

OpenBenchmarking.orgSeconds, Fewer Is BetterParboil 2.5Test: OpenMP MRI GriddingGCC 9.2.1GCC 10.0.0 201912081428425670SE +/- 0.15, N = 3SE +/- 0.24, N = 349.0062.941. (CXX) g++ options: -lm -lpthread -lgomp -O3 -ffast-math -fopenmp

CppPerformanceBenchmarks

Test: Random Numbers

OpenBenchmarking.orgSeconds, Fewer Is BetterCppPerformanceBenchmarks 9Test: Random NumbersGCC 10.0.0 20191208GCC 9.2.1LLVM Clang 10.0 20191222LLVM Clang 9.0.030060090012001500SE +/- 10.04, N = 3SE +/- 2.69, N = 3SE +/- 12.43, N = 3SE +/- 0.16, N = 31061.481069.481282.031296.981. (CXX) g++ options: -O3 -march=native -std=c++11

Tungsten Renderer

Scene: Non-Exponential

OpenBenchmarking.orgSeconds, Fewer Is BetterTungsten Renderer 0.2.2Scene: Non-ExponentialGCC 10.0.0 20191208GCC 9.2.1LLVM Clang 10.0 20191222LLVM Clang 9.0.00.91581.83162.74743.66324.579SE +/- 0.01371, N = 3SE +/- 0.01280, N = 3SE +/- 0.01369, N = 3SE +/- 0.00501, N = 33.369273.374194.051764.07043-fstrict-aliasing-fstrict-aliasing1. (CXX) g++ options: -O3 -march=native -std=c++0x -march=znver1 -msse2 -msse3 -mssse3 -msse4.1 -msse4.2 -msse4a -mfma -mbmi2 -mno-avx -mno-avx2 -mno-xop -mno-fma4 -mno-avx512f -mno-avx512vl -mno-avx512pf -mno-avx512er -mno-avx512cd -mno-avx512dq -mno-avx512bw -mno-avx512ifma -mno-avx512vbmi -rdynamic -ljpeg -lpthread -ldl

QMCPACK

OpenBenchmarking.orgTotal Execution Time - Seconds, Fewer Is BetterQMCPACK 3.8GCC 10.0.0 20191208GCC 9.2.1LLVM Clang 9.0.050010001500200025001878.11893.62262.8-finline-limit=1000 -funroll-all-loops-finline-limit=1000 -funroll-all-loops1. (CXX) g++ options: -O3 -march=native -fopenmp -fomit-frame-pointer -fstrict-aliasing -ffast-math -lm

FFTW

Build: Float + SSE - Size: 2D FFT Size 4096

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Float + SSE - Size: 2D FFT Size 4096GCC 9.2.1LLVM Clang 9.0.0GCC 10.0.0 201912085K10K15K20K25KSE +/- 309.06, N = 4SE +/- 404.65, N = 3SE +/- 239.62, N = 32418423559202871. (CC) gcc options: -pthread -O3 -march=native -lm

LAME MP3 Encoding

WAV To MP3

OpenBenchmarking.orgSeconds, Fewer Is BetterLAME MP3 Encoding 3.100WAV To MP3GCC 10.0.0 20191208GCC 9.2.1LLVM Clang 10.0 20191222LLVM Clang 9.0.0246810SE +/- 0.007, N = 3SE +/- 0.012, N = 3SE +/- 0.008, N = 3SE +/- 0.001, N = 66.7016.7297.7887.951-ffast-math -funroll-loops -fschedule-insns2 -fbranch-count-reg -fforce-addr-ffast-math -funroll-loops -fschedule-insns2 -fbranch-count-reg -fforce-addr-lncurses1. (CC) gcc options: -O3 -pipe -march=native -lm

CppPerformanceBenchmarks

Test: Stepanov Vector

OpenBenchmarking.orgSeconds, Fewer Is BetterCppPerformanceBenchmarks 9Test: Stepanov VectorLLVM Clang 10.0 20191222LLVM Clang 9.0.0GCC 10.0.0 20191208GCC 9.2.120406080100SE +/- 0.04, N = 3SE +/- 0.01, N = 3SE +/- 0.05, N = 3SE +/- 0.03, N = 366.7867.1176.2677.501. (CXX) g++ options: -O3 -march=native -std=c++11

AOBench

Size: 2048 x 2048 - Total Time

OpenBenchmarking.orgSeconds, Fewer Is BetterAOBenchSize: 2048 x 2048 - Total TimeGCC 10.0.0 20191208GCC 9.2.1LLVM Clang 9.0.0LLVM Clang 10.0 20191222714212835SE +/- 0.04, N = 3SE +/- 0.30, N = 3SE +/- 0.01, N = 3SE +/- 0.45, N = 1527.3928.6231.5931.751. (CC) gcc options: -lm -O3 -march=native

TSCP

AI Chess Performance

OpenBenchmarking.orgNodes Per Second, More Is BetterTSCP 1.81AI Chess PerformanceLLVM Clang 10.0 20191222LLVM Clang 9.0.0GCC 9.2.1GCC 10.0.0 20191208300K600K900K1200K1500KSE +/- 1690.40, N = 5SE +/- 2711.28, N = 5SE +/- 9469.71, N = 5SE +/- 1231.70, N = 515282481487001137308913459911. (CC) gcc options: -O3 -march=native

libgav1

Video Input: Summer Nature 4K

OpenBenchmarking.orgFPS, More Is Betterlibgav1 2019-10-05Video Input: Summer Nature 4KLLVM Clang 9.0.0GCC 10.0.0 20191208GCC 9.2.1LLVM Clang 10.0 20191222612182430SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.05, N = 323.9423.8923.7421.171. (CXX) g++ options: -O3 -march=native -lpthread

FLAC Audio Encoding

WAV To FLAC

OpenBenchmarking.orgSeconds, Fewer Is BetterFLAC Audio Encoding 1.3.2WAV To FLACLLVM Clang 10.0 20191222LLVM Clang 9.0.0GCC 10.0.0 20191208GCC 9.2.1246810SE +/- 0.003, N = 5SE +/- 0.005, N = 5SE +/- 0.006, N = 5SE +/- 0.007, N = 57.1457.1928.0428.073-fvisibility=hidden-fvisibility=hidden1. (CXX) g++ options: -O3 -march=native -logg -lm

libgav1

Video Input: Chimera 1080p

OpenBenchmarking.orgFPS, More Is Betterlibgav1 2019-10-05Video Input: Chimera 1080pLLVM Clang 9.0.0GCC 9.2.1GCC 10.0.0 20191208LLVM Clang 10.0 201912221224364860SE +/- 0.11, N = 3SE +/- 0.11, N = 3SE +/- 0.08, N = 3SE +/- 0.03, N = 352.4751.8451.7046.691. (CXX) g++ options: -O3 -march=native -lpthread

libgav1

Video Input: Summer Nature 1080p

OpenBenchmarking.orgFPS, More Is Betterlibgav1 2019-10-05Video Input: Summer Nature 1080pLLVM Clang 9.0.0GCC 10.0.0 20191208GCC 9.2.1LLVM Clang 10.0 2019122220406080100SE +/- 0.22, N = 3SE +/- 0.09, N = 3SE +/- 0.04, N = 3SE +/- 0.10, N = 378.5777.9277.1170.281. (CXX) g++ options: -O3 -march=native -lpthread

Tungsten Renderer

Scene: Hair

OpenBenchmarking.orgSeconds, Fewer Is BetterTungsten Renderer 0.2.2Scene: HairLLVM Clang 9.0.0LLVM Clang 10.0 20191222GCC 10.0.0 20191208GCC 9.2.13691215SE +/- 0.01343, N = 3SE +/- 0.01435, N = 3SE +/- 0.03984, N = 3SE +/- 0.02093, N = 38.700248.726259.449609.70952-fstrict-aliasing-fstrict-aliasing1. (CXX) g++ options: -O3 -march=native -std=c++0x -march=znver1 -msse2 -msse3 -mssse3 -msse4.1 -msse4.2 -msse4a -mfma -mbmi2 -mno-avx -mno-avx2 -mno-xop -mno-fma4 -mno-avx512f -mno-avx512vl -mno-avx512pf -mno-avx512er -mno-avx512cd -mno-avx512dq -mno-avx512bw -mno-avx512ifma -mno-avx512vbmi -rdynamic -ljpeg -lpthread -ldl

MKL-DNN DNNL

Harness: IP Batch 1D - Data Type: f32

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN DNNL 1.1Harness: IP Batch 1D - Data Type: f32LLVM Clang 9.0.0GCC 9.2.1GCC 10.0.0 201912080.41880.83761.25641.67522.094SE +/- 0.00428, N = 3SE +/- 0.00269, N = 3SE +/- 0.00636, N = 31.673151.737101.86131-fopenmp=libomp - MIN: 1.62-fopenmp - MIN: 1.67-fopenmp - MIN: 1.811. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -pie -lpthread -ldl

CppPerformanceBenchmarks

Test: Ctype

OpenBenchmarking.orgSeconds, Fewer Is BetterCppPerformanceBenchmarks 9Test: CtypeLLVM Clang 9.0.0LLVM Clang 10.0 20191222GCC 9.2.1GCC 10.0.0 20191208816243240SE +/- 0.01, N = 3SE +/- 0.30, N = 3SE +/- 0.00, N = 3SE +/- 0.04, N = 329.0929.8032.2232.311. (CXX) g++ options: -O3 -march=native -std=c++11

Tungsten Renderer

Scene: Volumetric Caustic

OpenBenchmarking.orgSeconds, Fewer Is BetterTungsten Renderer 0.2.2Scene: Volumetric CausticLLVM Clang 9.0.0LLVM Clang 10.0 20191222GCC 10.0.0 20191208GCC 9.2.10.91781.83562.75343.67124.589SE +/- 0.00668, N = 3SE +/- 0.00780, N = 3SE +/- 0.04581, N = 3SE +/- 0.03613, N = 33.681863.724883.966974.07902-fstrict-aliasing-fstrict-aliasing1. (CXX) g++ options: -O3 -march=native -std=c++0x -march=znver1 -msse2 -msse3 -mssse3 -msse4.1 -msse4.2 -msse4a -mfma -mbmi2 -mno-avx -mno-avx2 -mno-xop -mno-fma4 -mno-avx512f -mno-avx512vl -mno-avx512pf -mno-avx512er -mno-avx512cd -mno-avx512dq -mno-avx512bw -mno-avx512ifma -mno-avx512vbmi -rdynamic -ljpeg -lpthread -ldl

libgav1

Video Input: Chimera 1080p 10-bit

OpenBenchmarking.orgFPS, More Is Betterlibgav1 2019-10-05Video Input: Chimera 1080p 10-bitLLVM Clang 9.0.0LLVM Clang 10.0 20191222GCC 9.2.1GCC 10.0.0 20191208612182430SE +/- 0.09, N = 3SE +/- 0.07, N = 3SE +/- 0.04, N = 3SE +/- 0.04, N = 324.1923.1122.0121.931. (CXX) g++ options: -O3 -march=native -lpthread

CppPerformanceBenchmarks

Test: Stepanov Abstraction

OpenBenchmarking.orgSeconds, Fewer Is BetterCppPerformanceBenchmarks 9Test: Stepanov AbstractionLLVM Clang 10.0 20191222LLVM Clang 9.0.0GCC 10.0.0 20191208GCC 9.2.1714212835SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.05, N = 325.8626.0328.3028.521. (CXX) g++ options: -O3 -march=native -std=c++11

ASKAP

Test: tConvolve OpenMP - Degridding

OpenBenchmarking.orgMillion Grid Points Per Second, More Is BetterASKAP 2018-11-10Test: tConvolve OpenMP - DegriddingGCC 9.2.1GCC 10.0.0 201912089001800270036004500SE +/- 0.00, N = 3SE +/- 46.05, N = 34096.253716.331. (CXX) g++ options: -lpthread

Facebook RocksDB

Test: Random Fill

OpenBenchmarking.orgOp/s, More Is BetterFacebook RocksDB 6.3.6Test: Random FillGCC 9.2.1GCC 10.0.0 20191208200K400K600K800K1000KSE +/- 4040.85, N = 3SE +/- 13468.75, N = 310233359308971. (CXX) g++ options: -O3 -march=native -std=c++11 -fno-builtin-memcmp -fno-rtti -rdynamic -lpthread

Timed MrBayes Analysis

Primate Phylogeny Analysis

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed MrBayes Analysis 3.2.7Primate Phylogeny AnalysisLLVM Clang 9.0.0GCC 10.0.0 20191208GCC 9.2.1LLVM Clang 10.0 201912221632486480SE +/- 0.36, N = 3SE +/- 0.32, N = 3SE +/- 0.27, N = 3SE +/- 0.32, N = 364.8569.7970.3370.71-mabm-mabm1. (CC) gcc options: -mmmx -msse -msse2 -msse3 -mssse3 -msse4.1 -msse4.2 -msse4a -msha -maes -mavx -mfma -mavx2 -mrdrnd -mbmi -mbmi2 -madx -O3 -std=c99 -pedantic -march=native -lm

Himeno Benchmark

Poisson Pressure Solver

OpenBenchmarking.orgMFLOPS, More Is BetterHimeno Benchmark 3.0Poisson Pressure SolverLLVM Clang 9.0.0GCC 10.0.0 20191208LLVM Clang 10.0 20191222GCC 9.2.111002200330044005500SE +/- 84.96, N = 3SE +/- 51.56, N = 7SE +/- 77.03, N = 3SE +/- 61.37, N = 34983.284898.174866.904583.351. (CC) gcc options: -O3 -march=native -mavx2

FFTW

Build: Stock - Size: 2D FFT Size 4096

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Stock - Size: 2D FFT Size 4096GCC 10.0.0 20191208GCC 9.2.1LLVM Clang 9.0.02K4K6K8K10KSE +/- 76.26, N = 3SE +/- 26.44, N = 3SE +/- 24.98, N = 38135.28111.67531.21. (CC) gcc options: -pthread -O3 -march=native -lm

AOM AV1

AV1 Video Encoding

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 2019-09-16AV1 Video EncodingLLVM Clang 9.0.0LLVM Clang 10.0 20191222GCC 10.0.0 20191208GCC 9.2.10.03380.06760.10140.13520.169SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.150.150.150.141. (CXX) g++ options: -O3 -march=native -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread

CppPerformanceBenchmarks

Test: Function Objects

OpenBenchmarking.orgSeconds, Fewer Is BetterCppPerformanceBenchmarks 9Test: Function ObjectsGCC 10.0.0 20191208LLVM Clang 9.0.0LLVM Clang 10.0 20191222GCC 9.2.148121620SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.15, N = 313.8414.7014.7614.821. (CXX) g++ options: -O3 -march=native -std=c++11

SQLite Speedtest

Timed Time - Size 1,000

OpenBenchmarking.orgSeconds, Fewer Is BetterSQLite Speedtest 3.30Timed Time - Size 1,000GCC 9.2.1GCC 10.0.0 20191208LLVM Clang 9.0.0LLVM Clang 10.0 201912221326395265SE +/- 0.46, N = 3SE +/- 0.10, N = 3SE +/- 0.44, N = 3SE +/- 0.13, N = 356.1757.1258.3359.911. (CC) gcc options: -O3 -march=native -ldl -lz -lpthread

dav1d

Video Input: Chimera 1080p

OpenBenchmarking.orgFPS, More Is Betterdav1d 0.5.0Video Input: Chimera 1080pLLVM Clang 10.0 20191222LLVM Clang 9.0.0GCC 10.0.0 20191208GCC 9.2.1130260390520650SE +/- 1.62, N = 3SE +/- 2.49, N = 3SE +/- 4.43, N = 3SE +/- 2.39, N = 3625.11623.05612.94587.24MIN: 468.04 / MAX: 781MIN: 475.37 / MAX: 782.95MIN: 452.17 / MAX: 769.81MIN: 439.81 / MAX: 722.491. (CC) gcc options: -O3 -march=native -pthread

Tungsten Renderer

Scene: Water Caustic

OpenBenchmarking.orgSeconds, Fewer Is BetterTungsten Renderer 0.2.2Scene: Water CausticGCC 10.0.0 20191208LLVM Clang 10.0 20191222GCC 9.2.1LLVM Clang 9.0.0510152025SE +/- 0.03, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 318.2219.1619.2019.35-fstrict-aliasing-fstrict-aliasing1. (CXX) g++ options: -O3 -march=native -std=c++0x -march=znver1 -msse2 -msse3 -mssse3 -msse4.1 -msse4.2 -msse4a -mfma -mbmi2 -mno-avx -mno-avx2 -mno-xop -mno-fma4 -mno-avx512f -mno-avx512vl -mno-avx512pf -mno-avx512er -mno-avx512cd -mno-avx512dq -mno-avx512bw -mno-avx512ifma -mno-avx512vbmi -rdynamic -ljpeg -lpthread -ldl

VP9 libvpx Encoding

vpxenc VP9 1080p Video Encode

OpenBenchmarking.orgFrames Per Second, More Is BetterVP9 libvpx Encoding 1.8.1vpxenc VP9 1080p Video EncodeLLVM Clang 9.0.0LLVM Clang 10.0 20191222GCC 9.2.1GCC 10.0.0 2019120850100150200250SE +/- 1.20, N = 3SE +/- 1.73, N = 15SE +/- 1.49, N = 3SE +/- 1.29, N = 3207.78207.68197.37195.631. (CXX) g++ options: -m64 -lm -lpthread -O3 -march=native -fPIC -U_FORTIFY_SOURCE -std=c++11

Stockfish

Total Time

OpenBenchmarking.orgNodes Per Second, More Is BetterStockfish 9Total TimeLLVM Clang 9.0.0GCC 10.0.0 20191208GCC 9.2.120M40M60M80M100MSE +/- 194228.81, N = 3SE +/- 1221575.84, N = 3SE +/- 74252.03, N = 38133605480909082767374621. (CXX) g++ options: -m64 -lpthread -O3 -march=native -fno-exceptions -std=c++11 -pedantic -msse -msse3 -mpopcnt -flto

Facebook RocksDB

Test: Sequential Fill

OpenBenchmarking.orgOp/s, More Is BetterFacebook RocksDB 6.3.6Test: Sequential FillGCC 9.2.1GCC 10.0.0 20191208200K400K600K800K1000KSE +/- 10427.62, N = 3SE +/- 4276.33, N = 3108364910247331. (CXX) g++ options: -O3 -march=native -std=c++11 -fno-builtin-memcmp -fno-rtti -rdynamic -lpthread

LeelaChessZero

Backend: Random

OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.22.0Backend: RandomGCC 9.2.1GCC 10.0.0 20191208LLVM Clang 9.0.020K40K60K80K100KSE +/- 354.64, N = 3SE +/- 275.83, N = 3SE +/- 553.66, N = 3105996.0105805.0100236.81. (CXX) g++ options: -O3 -march=native -lpthread

Rodinia

Test: OpenMP CFD Solver

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 2.4Test: OpenMP CFD SolverGCC 9.2.1GCC 10.0.0 20191208LLVM Clang 9.0.03691215SE +/- 0.052, N = 3SE +/- 0.029, N = 3SE +/- 0.133, N = 39.1519.1859.633-O2 -lOpenCL-O2 -lOpenCL-O3 -fopenmp1. (CXX) g++ options:

MKL-DNN DNNL

Harness: Convolution Batch conv_alexnet - Data Type: f32

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN DNNL 1.1Harness: Convolution Batch conv_alexnet - Data Type: f32LLVM Clang 9.0.0GCC 10.0.0 20191208GCC 9.2.1306090120150SE +/- 0.15, N = 3SE +/- 1.45, N = 3SE +/- 1.55, N = 3118.99124.28125.01-fopenmp=libomp - MIN: 118.28-fopenmp - MIN: 122.25-fopenmp - MIN: 122.481. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -pie -lpthread -ldl

ASKAP

Test: tConvolve OpenMP - Gridding

OpenBenchmarking.orgMillion Grid Points Per Second, More Is BetterASKAP 2018-11-10Test: tConvolve OpenMP - GriddingGCC 9.2.1GCC 10.0.0 2019120812002400360048006000SE +/- 37.73, N = 3SE +/- 34.80, N = 35509.305255.511. (CXX) g++ options: -lpthread

CppPerformanceBenchmarks

Test: Math Library

OpenBenchmarking.orgSeconds, Fewer Is BetterCppPerformanceBenchmarks 9Test: Math LibraryLLVM Clang 10.0 20191222LLVM Clang 9.0.0GCC 10.0.0 20191208GCC 9.2.160120180240300SE +/- 2.46, N = 3SE +/- 0.41, N = 3SE +/- 0.29, N = 3SE +/- 0.45, N = 3258.40260.64262.97269.481. (CXX) g++ options: -O3 -march=native -std=c++11

x265

H.265 1080p Video Encoding

OpenBenchmarking.orgFrames Per Second, More Is Betterx265 3.1.2H.265 1080p Video EncodingLLVM Clang 10.0 20191222LLVM Clang 9.0.0GCC 10.0.0 20191208GCC 9.2.11530456075SE +/- 0.39, N = 3SE +/- 0.07, N = 3SE +/- 0.29, N = 3SE +/- 0.10, N = 367.1166.2765.7464.691. (CXX) g++ options: -O3 -march=native -rdynamic -lpthread -lrt -ldl -lnuma

miniFE

Problem Size: Small

OpenBenchmarking.orgCG Mflops, More Is BetterminiFE 2.2Problem Size: SmallLLVM Clang 9.0.0GCC 9.2.1GCC 10.0.0 201912082K4K6K8K10KSE +/- 3.20, N = 3SE +/- 9.20, N = 3SE +/- 4.74, N = 38037.337777.507767.981. (CXX) g++ options: -O3 -fopenmp -pthread -lmpi_cxx -lmpi

PostgreSQL pgbench

Scaling: Buffer Test - Test: Normal Load - Mode: Read Only

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 12.0Scaling: Buffer Test - Test: Normal Load - Mode: Read OnlyLLVM Clang 9.0.0GCC 10.0.0 20191208LLVM Clang 10.0 20191222GCC 9.2.1140K280K420K560K700KSE +/- 2145.94, N = 3SE +/- 876.37, N = 3SE +/- 1018.99, N = 3SE +/- 1703.60, N = 3675821.10673644.96668836.61654500.681. (CC) gcc options: -fno-strict-aliasing -fwrapv -O3 -march=native -lpgcommon -lpgport -lpq -lpthread -lrt -lcrypt -ldl -lm

Zstd Compression

Compressing ubuntu-16.04.3-server-i386.img, Compression Level 19

OpenBenchmarking.orgSeconds, Fewer Is BetterZstd Compression 1.3.4Compressing ubuntu-16.04.3-server-i386.img, Compression Level 19LLVM Clang 9.0.0GCC 9.2.1GCC 10.0.0 20191208LLVM Clang 10.0 201912223691215SE +/- 0.100, N = 3SE +/- 0.125, N = 3SE +/- 0.089, N = 3SE +/- 0.033, N = 39.81710.04010.05210.1071. (CC) gcc options: -O3 -march=native -pthread -lz -llzma

dav1d

Video Input: Summer Nature 1080p

OpenBenchmarking.orgFPS, More Is Betterdav1d 0.5.0Video Input: Summer Nature 1080pLLVM Clang 10.0 20191222GCC 10.0.0 20191208LLVM Clang 9.0.0GCC 9.2.1150300450600750SE +/- 2.78, N = 3SE +/- 0.42, N = 3SE +/- 1.73, N = 3SE +/- 1.97, N = 3685.40676.37674.39667.33MIN: 400.43 / MAX: 750.08MIN: 396.97 / MAX: 738.73MIN: 376.28 / MAX: 738.32MIN: 387.89 / MAX: 728.981. (CC) gcc options: -O3 -march=native -pthread

dav1d

Video Input: Summer Nature 4K

OpenBenchmarking.orgFPS, More Is Betterdav1d 0.5.0Video Input: Summer Nature 4KLLVM Clang 10.0 20191222GCC 10.0.0 20191208LLVM Clang 9.0.0GCC 9.2.160120180240300SE +/- 0.34, N = 3SE +/- 0.64, N = 3SE +/- 0.37, N = 3SE +/- 1.13, N = 3291.76289.29288.60285.28MIN: 174.15 / MAX: 309.6MIN: 172.47 / MAX: 306.73MIN: 170.73 / MAX: 306.52MIN: 168.88 / MAX: 304.291. (CC) gcc options: -O3 -march=native -pthread

MKL-DNN DNNL

Harness: Convolution Batch conv_googlenet_v3 - Data Type: f32

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN DNNL 1.1Harness: Convolution Batch conv_googlenet_v3 - Data Type: f32LLVM Clang 9.0.0GCC 9.2.1GCC 10.0.0 201912081224364860SE +/- 0.11, N = 3SE +/- 0.18, N = 3SE +/- 0.24, N = 351.3352.2852.47-fopenmp=libomp - MIN: 50.48-fopenmp - MIN: 51.18-fopenmp - MIN: 51.431. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -pie -lpthread -ldl

Facebook RocksDB

Test: Random Read

OpenBenchmarking.orgOp/s, More Is BetterFacebook RocksDB 6.3.6Test: Random ReadGCC 10.0.0 20191208GCC 9.2.130M60M90M120M150MSE +/- 467553.12, N = 3SE +/- 234281.58, N = 31453689671422877661. (CXX) g++ options: -O3 -march=native -std=c++11 -fno-builtin-memcmp -fno-rtti -rdynamic -lpthread

BYTE Unix Benchmark

Computational Test: Dhrystone 2

OpenBenchmarking.orgLPS, More Is BetterBYTE Unix Benchmark 3.6Computational Test: Dhrystone 2GCC 9.2.1GCC 10.0.0 2019120810M20M30M40M50MSE +/- 70447.90, N = 3SE +/- 638551.10, N = 448969411.248002215.11. (CC) gcc options: -O3 -march=native

XZ Compression

Compressing ubuntu-16.04.3-server-i386.img, Compression Level 9

OpenBenchmarking.orgSeconds, Fewer Is BetterXZ Compression 5.2.4Compressing ubuntu-16.04.3-server-i386.img, Compression Level 9LLVM Clang 9.0.0LLVM Clang 10.0 20191222GCC 9.2.1GCC 10.0.0 20191208510152025SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.16, N = 319.6919.6919.7520.041. (CC) gcc options: -pthread -fvisibility=hidden -O3 -march=native

NGINX Benchmark

Static Web Page Serving

OpenBenchmarking.orgRequests Per Second, More Is BetterNGINX Benchmark 1.9.9Static Web Page ServingLLVM Clang 9.0.0LLVM Clang 10.0 20191222GCC 9.2.1GCC 10.0.0 201912089K18K27K36K45KSE +/- 326.86, N = 3SE +/- 238.62, N = 3SE +/- 457.53, N = 3SE +/- 490.20, N = 343783.3043580.6043426.7243035.461. (CC) gcc options: -lpthread -lcrypt -lcrypto -lz -O3 -march=native

x264

H.264 Video Encoding

OpenBenchmarking.orgFrames Per Second, More Is Betterx264 2018-09-25H.264 Video EncodingGCC 10.0.0 20191208LLVM Clang 9.0.0GCC 9.2.1LLVM Clang 10.0 201912224080120160200SE +/- 2.04, N = 8SE +/- 1.58, N = 12SE +/- 0.53, N = 3SE +/- 0.76, N = 3199.22198.29197.93196.10-mstack-alignment=64-mstack-alignment=641. (CC) gcc options: -ldl -m64 -lm -lpthread -O3 -ffast-math -march=native -std=gnu99 -fPIC -fomit-frame-pointer -fno-tree-vectorize

CppPerformanceBenchmarks

Test: Atol

OpenBenchmarking.orgSeconds, Fewer Is BetterCppPerformanceBenchmarks 9Test: AtolGCC 9.2.1LLVM Clang 10.0 20191222LLVM Clang 9.0.0GCC 10.0.0 201912081326395265SE +/- 0.20, N = 3SE +/- 0.26, N = 3SE +/- 0.25, N = 3SE +/- 0.27, N = 357.4757.8458.0058.281. (CXX) g++ options: -O3 -march=native -std=c++11

N-Queens

Elapsed Time

OpenBenchmarking.orgSeconds, Fewer Is BetterN-Queens 1.0Elapsed TimeGCC 10.0.0 20191208GCC 9.2.10.99651.9932.98953.9864.9825SE +/- 0.004, N = 3SE +/- 0.007, N = 34.3784.4291. (CC) gcc options: -static -fopenmp -O3 -march=native

ASKAP

Test: tConvolve MT - Degridding

OpenBenchmarking.orgMillion Grid Points Per Second, More Is BetterASKAP 2018-11-10Test: tConvolve MT - DegriddingGCC 9.2.1GCC 10.0.0 201912087001400210028003500SE +/- 2.38, N = 3SE +/- 13.81, N = 33376.273339.331. (CXX) g++ options: -lpthread

Crafty

Elapsed Time

OpenBenchmarking.orgNodes Per Second, More Is BetterCrafty 25.2Elapsed TimeGCC 10.0.0 20191208GCC 9.2.12M4M6M8M10MSE +/- 15301.93, N = 3SE +/- 11113.29, N = 3902783589599371. (CC) gcc options: -pthread -lstdc++ -fprofile-use -lm

ASKAP

Test: tConvolve MT - Gridding

OpenBenchmarking.orgMillion Grid Points Per Second, More Is BetterASKAP 2018-11-10Test: tConvolve MT - GriddingGCC 9.2.1GCC 10.0.0 20191208400800120016002000SE +/- 1.51, N = 3SE +/- 10.21, N = 31955.971943.581. (CXX) g++ options: -lpthread

Apache Benchmark

Static Web Page Serving

OpenBenchmarking.orgRequests Per Second, More Is BetterApache Benchmark 2.4.29Static Web Page ServingLLVM Clang 10.0 20191222GCC 9.2.1GCC 10.0.0 20191208LLVM Clang 9.0.07K14K21K28K35KSE +/- 71.82, N = 3SE +/- 27.04, N = 3SE +/- 54.98, N = 3SE +/- 32.67, N = 334327.9034228.5034157.7634144.391. (CC) gcc options: -shared -fPIC -pthread -O3 -march=native

Facebook RocksDB

Test: Random Fill Sync

OpenBenchmarking.orgOp/s, More Is BetterFacebook RocksDB 6.3.6Test: Random Fill SyncGCC 9.2.1GCC 10.0.0 201912085K10K15K20K25KSE +/- 29.29, N = 3SE +/- 40.43, N = 324368242771. (CXX) g++ options: -O3 -march=native -std=c++11 -fno-builtin-memcmp -fno-rtti -rdynamic -lpthread

SQLite

Threads / Copies: 1

OpenBenchmarking.orgSeconds, Fewer Is BetterSQLite 3.30.1Threads / Copies: 1LLVM Clang 9.0.0GCC 10.0.0 20191208LLVM Clang 10.0 20191222GCC 9.2.148121620SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 314.2014.2414.2514.251. (CC) gcc options: -O3 -march=native -lz -lm -ldl -lpthread

Parboil

Test: OpenMP Stencil

OpenBenchmarking.orgSeconds, Fewer Is BetterParboil 2.5Test: OpenMP StencilGCC 10.0.0 20191208GCC 9.2.1246810SE +/- 0.036635, N = 3SE +/- 0.021046, N = 37.5517067.5763961. (CXX) g++ options: -lm -lpthread -lgomp -O3 -ffast-math -fopenmp

Rodinia

Test: OpenMP LavaMD

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 2.4Test: OpenMP LavaMDGCC 10.0.0 20191208GCC 9.2.13691215SE +/- 0.01, N = 3SE +/- 0.02, N = 310.0910.101. (CXX) g++ options: -O2 -lOpenCL

Facebook RocksDB

Test: Read While Writing

OpenBenchmarking.orgOp/s, More Is BetterFacebook RocksDB 6.3.6Test: Read While WritingGCC 9.2.1GCC 10.0.0 201912081.1M2.2M3.3M4.4M5.5MSE +/- 73789.66, N = 4SE +/- 38218.40, N = 15495089049459161. (CXX) g++ options: -O3 -march=native -std=c++11 -fno-builtin-memcmp -fno-rtti -rdynamic -lpthread

GROMACS

Water Benchmark

OpenBenchmarking.orgNs Per Day, More Is BetterGROMACS 2019.4Water BenchmarkGCC 9.2.1GCC 10.0.0 201912080.56611.13221.69832.26442.8305SE +/- 0.001, N = 3SE +/- 0.005, N = 32.5162.5151. (CXX) g++ options: -mavx2 -mfma -O3 -march=native -std=c++11 -funroll-all-loops -pthread -lrt -lpthread -lm

Parboil

Test: OpenMP CUTCP

OpenBenchmarking.orgSeconds, Fewer Is BetterParboil 2.5Test: OpenMP CUTCPGCC 10.0.0 20191208GCC 9.2.10.2830.5660.8491.1321.415SE +/- 0.007993, N = 3SE +/- 0.002849, N = 31.2574101.2577161. (CXX) g++ options: -lm -lpthread -lgomp -O3 -ffast-math -fopenmp

PostgreSQL pgbench

Scaling: Buffer Test - Test: Normal Load - Mode: Read Write

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 12.0Scaling: Buffer Test - Test: Normal Load - Mode: Read WriteLLVM Clang 9.0.0GCC 9.2.1GCC 10.0.0 20191208LLVM Clang 10.0 201912227K14K21K28K35KSE +/- 127.72, N = 3SE +/- 77.46, N = 3SE +/- 2238.03, N = 12SE +/- 1738.68, N = 1531491.3830910.9915513.4414868.681. (CC) gcc options: -fno-strict-aliasing -fwrapv -O3 -march=native -lpgcommon -lpgport -lpq -lpthread -lrt -lcrypt -ldl -lm

Smallpt

Global Illumination Renderer; 128 Samples

OpenBenchmarking.orgSeconds, Fewer Is BetterSmallpt 1.0Global Illumination Renderer; 128 SamplesGCC 9.2.1GCC 10.0.0 201912081.07442.14883.22324.29765.372SE +/- 0.006, N = 3SE +/- 0.107, N = 153.4924.7751. (CXX) g++ options: -fopenmp -O3 -march=native

Rodinia

Test: OpenMP Streamcluster

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 2.4Test: OpenMP StreamclusterLLVM Clang 9.0.0GCC 10.0.0 20191208GCC 9.2.1510152025SE +/- 0.06, N = 3SE +/- 0.02, N = 3SE +/- 0.54, N = 1518.8219.0719.76-O3 -fopenmp-O2 -lOpenCL-O2 -lOpenCL1. (CXX) g++ options:

LeelaChessZero

Backend: BLAS

OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.22.0Backend: BLASLLVM Clang 9.0.0GCC 9.2.1GCC 10.0.0 20191208816243240SE +/- 0.60, N = 15SE +/- 0.56, N = 12SE +/- 0.41, N = 333.3032.7629.331. (CXX) g++ options: -O3 -march=native -lpthread

Phoronix Test Suite v10.8.5