AARCH64 codegen comparison update

gcc7's performance on Cortex A53 (32kB L1)

HTML result view exported from: https://openbenchmarking.org/result/1703238-RI-GCCLATEST31&sro.

AARCH64 codegen comparison updateProcessorMotherboardMemoryDiskOSKernelCompilerFile-SystemScreen ResolutionA53 vectorize, pre-patchthunderx/vectorize, pre-patchA53 vectorize/LTO, pre patchA53, post patchA53 mtune/vectorize, post-patchA53 vectorize, updatedA53 vectorize, earlier buildA57 vectorize/unrolled GCC 7.0.1A53 vectorize GCC 7.0.1A57 vectorize/unrolled GCC 6.3AArch64 rev 4 @ 1.50GHz (4 Cores)Amlogic2048MB32GB 00000 + 16GB NCardUbuntu 16.043.14.29 (aarch64)GCC 7.0.0 20170110 + Clang 3.8.0-2ubuntu4 + LLVM 3.8.0ext41920x3240AArch64 rev 4 @ 1.55GHz (4 Cores)GCC 7.0.0 20170113 + Clang 3.8.0-2ubuntu4 + LLVM 3.8.0Unknown @ 1.54GHz (4 Cores)16GB NCard + 32GB 000003.14.79-vegas95 (aarch64)GCC 7.0.1 20170214 + Clang 3.8.0-2ubuntu4 + LLVM 3.8.01280x1440GCC 7.0.1 20170220 + Clang 3.8.0-2ubuntu4 + LLVM 3.8.08GB NCard + 32GB 00000GCC 7.0.1 20170322 + Clang 3.8.0-2ubuntu4 + LLVM 3.8.0GCC 6.3.1 20170316 + Clang 3.8.0-2ubuntu4 + LLVM 3.8.0OpenBenchmarking.orgCompiler Details- A53 vectorize, pre-patch: --build=aarch64-linux-gnu --disable-browser-plugin --disable-libquadmath --disable-werror --enable-checking=release --enable-clocale=gnu --enable-fix-cortex-a53-843419 --enable-gnu-unique-object --enable-languages=c,c++,fortran --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-plugin --enable-shared --enable-threads=posix --host=aarch64-linux-gnu --target=aarch64-linux-gnu --with-arch-directory=aarch64 --with-default-libstdcxx-abi=new- thunderx/vectorize, pre-patch: --build=aarch64-linux-gnu --disable-browser-plugin --disable-libquadmath --disable-werror --enable-checking=release --enable-clocale=gnu --enable-fix-cortex-a53-843419 --enable-gnu-unique-object --enable-languages=c,c++,fortran --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-plugin --enable-shared --enable-threads=posix --host=aarch64-linux-gnu --target=aarch64-linux-gnu --with-arch-directory=aarch64 --with-default-libstdcxx-abi=new- A53 vectorize/LTO, pre patch: --build=aarch64-linux-gnu --disable-browser-plugin --disable-libquadmath --disable-werror --enable-checking=release --enable-clocale=gnu --enable-fix-cortex-a53-843419 --enable-gnu-unique-object --enable-languages=c,c++,fortran --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-plugin --enable-shared --enable-threads=posix --host=aarch64-linux-gnu --target=aarch64-linux-gnu --with-arch-directory=aarch64 --with-default-libstdcxx-abi=new- A53, post patch: --build=aarch64-linux-gnu --disable-browser-plugin --disable-libquadmath --disable-werror --enable-checking=release --enable-clocale=gnu --enable-fix-cortex-a53-843419 --enable-gnu-unique-object --enable-languages=c,c++,fortran --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-plugin --enable-shared --enable-threads=posix --host=aarch64-linux-gnu --target=aarch64-linux-gnu --with-arch-directory=aarch64 --with-default-libstdcxx-abi=new- A53 mtune/vectorize, post-patch: --build=aarch64-linux-gnu --disable-browser-plugin --disable-libquadmath --disable-werror --enable-checking=release --enable-clocale=gnu --enable-fix-cortex-a53-843419 --enable-gnu-unique-object --enable-languages=c,c++,fortran --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-plugin --enable-shared --enable-threads=posix --host=aarch64-linux-gnu --target=aarch64-linux-gnu --with-arch-directory=aarch64 --with-default-libstdcxx-abi=new- A53 vectorize, updated: --build=aarch64-linux-gnu --disable-browser-plugin --disable-libquadmath --disable-werror --enable-checking=release --enable-clocale=gnu --enable-fix-cortex-a53-843419 --enable-gnu-unique-object --enable-languages=c,c++,fortran --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-plugin --enable-shared --enable-threads=posix --host=aarch64-linux-gnu --target=aarch64-linux-gnu --with-arch-directory=aarch64 --with-default-libstdcxx-abi=new- A53 vectorize, earlier build: --build=aarch64-linux-gnu --disable-browser-plugin --disable-libquadmath --disable-werror --enable-checking=release --enable-clocale=gnu --enable-fix-cortex-a53-843419 --enable-gnu-unique-object --enable-languages=c,c++,fortran --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-plugin --enable-shared --enable-threads=posix --host=aarch64-linux-gnu --target=aarch64-linux-gnu --with-arch-directory=aarch64 --with-default-libstdcxx-abi=new- A57 vectorize/unrolled GCC 7.0.1: --build=aarch64-linux-gnu --disable-bootstrap --disable-browser-plugin --disable-libquadmath --disable-werror --enable-checking=release --enable-clocale=gnu --enable-gnu-unique-object --enable-languages=c,c++,fortran --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-plugin --enable-shared --enable-threads=posix --host=aarch64-linux-gnu --target=aarch64-linux-gnu --with-arch-directory=aarch64 --with-default-libstdcxx-abi=new- A53 vectorize GCC 7.0.1: --build=aarch64-linux-gnu --disable-bootstrap --disable-browser-plugin --disable-libquadmath --disable-werror --enable-checking=release --enable-clocale=gnu --enable-gnu-unique-object --enable-languages=c,c++,fortran --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-plugin --enable-shared --enable-threads=posix --host=aarch64-linux-gnu --target=aarch64-linux-gnu --with-arch-directory=aarch64 --with-default-libstdcxx-abi=new- A57 vectorize/unrolled GCC 6.3: --build=aarch64-linux-gnu --disable-bootstrap --disable-browser-plugin --disable-libquadmath --disable-werror --enable-checking=release --enable-clocale=gnu --enable-gnu-unique-object --enable-languages=c,c++,fortran --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-plugin --enable-shared --enable-threads=posix --host=aarch64-linux-gnu --target=aarch64-linux-gnu --with-arch-directory=aarch64 --with-default-libstdcxx-abi=newDisk Details- A53 vectorize, pre-patch: CFQ / commit=30,errors=remount-ro,noatime,nodiratime,rw- thunderx/vectorize, pre-patch: CFQ / commit=30,errors=remount-ro,noatime,nodiratime,rw- A53 vectorize/LTO, pre patch: CFQ / commit=30,errors=remount-ro,noatime,nodiratime,rw- A53, post patch: CFQ / commit=30,errors=remount-ro,noatime,nodiratime,rw- A53 mtune/vectorize, post-patch: CFQ / commit=30,errors=remount-ro,noatime,nodiratime,rw- A53 vectorize, updated: DEADLINE / commit=45,errors=remount-ro,noatime,nodiratime,rw- A53 vectorize, earlier build: CFQ / commit=45,errors=remount-ro,noatime,nodiratime,rw- A57 vectorize/unrolled GCC 7.0.1: CFQ / commit=120,errors=remount-ro,noatime,nodiratime,rw- A53 vectorize GCC 7.0.1: CFQ / commit=120,errors=remount-ro,noatime,nodiratime,rw- A57 vectorize/unrolled GCC 6.3: CFQ / commit=120,errors=remount-ro,noatime,nodiratime,rwProcessor Details- A53 vectorize, pre-patch: Scaling Governor: meson_cpufreq performance- thunderx/vectorize, pre-patch: Scaling Governor: meson_cpufreq performance- A53 vectorize/LTO, pre patch: Scaling Governor: meson_cpufreq performance- A53, post patch: Scaling Governor: meson_cpufreq performance- A53 mtune/vectorize, post-patch: Scaling Governor: meson_cpufreq interactive- A53 vectorize, updated: Scaling Governor: meson_cpufreq performance- A53 vectorize, earlier build: Scaling Governor: meson_cpufreq performance- A57 vectorize/unrolled GCC 7.0.1: Scaling Governor: meson_cpufreq performance- A53 vectorize GCC 7.0.1: Scaling Governor: meson_cpufreq performance- A57 vectorize/unrolled GCC 6.3: Scaling Governor: meson_cpufreq performance

AARCH64 codegen comparison updatepostmark: Disk Transaction Performanceramspeed: Copy - Integerramspeed: Copy - Floating Pointfftw: Stock - 2D FFT Size 2048mafft: Multiple Sequence Alignmentgmpbench: Total Timefhourstones: Complex Connect-4 Solvingttsiod-renderer: Phong Rendering With Soft-Shadow Mappingc-ray: Total Timeprimesieve: 1e12 Prime Number Generationsmallpt: Global Illumination Renderer; 100 Samplessudokut: Total Timetachyon: Total Timeopenssl: RSA 4096-bit Performanceredis: GETA53 vectorize, pre-patchthunderx/vectorize, pre-patchA53 vectorize/LTO, pre patchA53, post patchA53 mtune/vectorize, post-patchA53 vectorize, updatedA53 vectorize, earlier buildA57 vectorize/unrolled GCC 7.0.1A53 vectorize GCC 7.0.1A57 vectorize/unrolled GCC 6.313634581.324580.39196.9035.42552.843212.1023.16187.97543.16167101.9569.2721.50310344.7313512821.432817.45190.6334.46554.833210.2023.01149.82566.21167102.7571.4121.50318926.0213784829.914825.13180.5333.16554.373213.7723.77184.81540.95168101.7567.6421.50311785.0213814965.064964.66186.2133.06552.563209.6723.47186.69553.13168101.8869.4021.50309030.6413784955.974965.60184.8132.17555.103205.4023.49186.61573.13167102.1769.3421.50313438.9112174706.404785.59185.1533.90554.113223.5723.29161.80574.65166102.7269.9021.40277268.2312114816.694816.71191.5434.52554.213233.7723.49162.23523.43166102.5969.3921.40283742.8311844201.534193.97173.0334.22553.173398.4723.56154.81525.00168103.0421.47276169.8611944161.094188.18156.7235.52552.753415.0723.13151.02547.12166102.9769.6421.47276298.4411904384.534388.73157.9735.47554.033325.6022.29149.61531.95169103.0072.2821.20275458.70OpenBenchmarking.org

PostMark

Disk Transaction Performance

OpenBenchmarking.orgTPS, More Is BetterPostMark 1.51Disk Transaction PerformanceA53 mtune/vectorize, post-patchA53 vectorize GCC 7.0.1A53 vectorize, earlier buildA53 vectorize, pre-patchA53 vectorize, updatedA53 vectorize/LTO, pre patchA53, post patchA57 vectorize/unrolled GCC 6.3A57 vectorize/unrolled GCC 7.0.1thunderx/vectorize, pre-patch30060090012001500SE +/- 5.00, N = 3SE +/- 3.67, N = 3SE +/- 5.29, N = 3SE +/- 2.67, N = 3SE +/- 2.00, N = 3SE +/- 2.67, N = 3SE +/- 4.33, N = 3SE +/- 0.00, N = 313781194121113631217137813811190118413511. (CC) gcc options: -O3

RAMspeed SMP

Type: Copy - Benchmark: Integer

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Type: Copy - Benchmark: IntegerA53 mtune/vectorize, post-patchA53 vectorize GCC 7.0.1A53 vectorize, earlier buildA53 vectorize, pre-patchA53 vectorize, updatedA53 vectorize/LTO, pre patchA53, post patchA57 vectorize/unrolled GCC 6.3A57 vectorize/unrolled GCC 7.0.1thunderx/vectorize, pre-patch110022003300440055004955.974161.094816.694581.324706.404829.914965.064384.534201.532821.43

RAMspeed SMP

Type: Copy - Benchmark: Floating Point

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Type: Copy - Benchmark: Floating PointA53 mtune/vectorize, post-patchA53 vectorize GCC 7.0.1A53 vectorize, earlier buildA53 vectorize, pre-patchA53 vectorize, updatedA53 vectorize/LTO, pre patchA53, post patchA57 vectorize/unrolled GCC 6.3A57 vectorize/unrolled GCC 7.0.1thunderx/vectorize, pre-patch110022003300440055004965.604188.184816.714580.394785.594825.134964.664388.734193.972817.45

FFTW

Build: Stock - Size: 2D FFT Size 2048

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.4Build: Stock - Size: 2D FFT Size 2048A53 mtune/vectorize, post-patchA53 vectorize GCC 7.0.1A53 vectorize, earlier buildA53 vectorize, pre-patchA53 vectorize, updatedA53 vectorize/LTO, pre patchA53, post patchA57 vectorize/unrolled GCC 6.3A57 vectorize/unrolled GCC 7.0.1thunderx/vectorize, pre-patch4080120160200SE +/- 0.21, N = 5SE +/- 0.16, N = 5SE +/- 0.16, N = 5SE +/- 0.99, N = 5SE +/- 0.06, N = 5SE +/- 0.49, N = 5SE +/- 0.08, N = 5SE +/- 0.26, N = 5SE +/- 0.10, N = 5SE +/- 1.10, N = 5184.81156.72191.54196.90185.15180.53186.21157.97173.03190.63-Ofast -mtune=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-O3 -mcpu=cortex-a53 -ftree-vectorize -flto -ffat-lto-objects-Ofast -mcpu=cortex-a53-Ofast -mtune=cortex-a57 -ftree-vectorize -funroll-loops-Ofast -mtune=cortex-a57 -ftree-vectorize -funroll-loops-Ofast -mcpu=thunderx -ftree-vectorize1. (CC) gcc options: -fomit-frame-pointer -fipa-pta -march=armv8-a+crc -lm

Timed MAFFT Alignment

Multiple Sequence Alignment

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed MAFFT Alignment 6.864Multiple Sequence AlignmentA53 mtune/vectorize, post-patchA53 vectorize GCC 7.0.1A53 vectorize, earlier buildA53 vectorize, pre-patchA53 vectorize, updatedA53 vectorize/LTO, pre patchA53, post patchA57 vectorize/unrolled GCC 6.3A57 vectorize/unrolled GCC 7.0.1thunderx/vectorize, pre-patch816243240SE +/- 0.01, N = 3SE +/- 0.97, N = 6SE +/- 1.04, N = 6SE +/- 0.80, N = 6SE +/- 0.79, N = 6SE +/- 0.70, N = 6SE +/- 0.71, N = 6SE +/- 0.08, N = 3SE +/- 0.71, N = 6SE +/- 0.73, N = 632.1735.5234.5235.4233.9033.1633.0635.4734.2234.461. (CC) gcc options: -O3 -lm -lpthread

GMPbench

Total Time

OpenBenchmarking.orgGMPbench Score, More Is BetterGMPbench 0.2Total TimeA53 mtune/vectorize, post-patchA53 vectorize GCC 7.0.1A53 vectorize, earlier buildA53 vectorize, pre-patchA53 vectorize, updatedA53 vectorize/LTO, pre patchA53, post patchA57 vectorize/unrolled GCC 6.3A57 vectorize/unrolled GCC 7.0.1thunderx/vectorize, pre-patch120240360480600555.10552.75554.21552.84554.11554.37552.56554.03553.17554.83-Ofast -mtune=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-O3 -mcpu=cortex-a53 -ftree-vectorize -flto -ffat-lto-objects-Ofast -mcpu=cortex-a53-Ofast -mtune=cortex-a57 -ftree-vectorize -funroll-loops-Ofast -mtune=cortex-a57 -ftree-vectorize -funroll-loops-Ofast -mcpu=thunderx -ftree-vectorize1. (CC) gcc options: -fomit-frame-pointer -fipa-pta -march=armv8-a+crc -lm

Fhourstones

Complex Connect-4 Solving

OpenBenchmarking.orgKpos / sec, More Is BetterFhourstones 3.1Complex Connect-4 SolvingA53 mtune/vectorize, post-patchA53 vectorize GCC 7.0.1A53 vectorize, earlier buildA53 vectorize, pre-patchA53 vectorize, updatedA53 vectorize/LTO, pre patchA53, post patchA57 vectorize/unrolled GCC 6.3A57 vectorize/unrolled GCC 7.0.1thunderx/vectorize, pre-patch7001400210028003500SE +/- 1.81, N = 3SE +/- 2.42, N = 3SE +/- 1.49, N = 3SE +/- 0.35, N = 3SE +/- 3.32, N = 3SE +/- 0.22, N = 3SE +/- 1.47, N = 3SE +/- 1.97, N = 3SE +/- 1.93, N = 3SE +/- 0.76, N = 33205.403415.073233.773212.103223.573213.773209.673325.603398.473210.201. (CC) gcc options: -O3

TTSIOD 3D Renderer

Phong Rendering With Soft-Shadow Mapping

OpenBenchmarking.orgFPS, More Is BetterTTSIOD 3D Renderer 2.3aPhong Rendering With Soft-Shadow MappingA53 mtune/vectorize, post-patchA53 vectorize GCC 7.0.1A53 vectorize, earlier buildA53 vectorize, pre-patchA53 vectorize, updatedA53 vectorize/LTO, pre patchA53, post patchA57 vectorize/unrolled GCC 6.3A57 vectorize/unrolled GCC 7.0.1thunderx/vectorize, pre-patch612182430SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.09, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 323.4923.1323.4923.1623.2923.7723.4722.2923.5623.01-Ofast -mtune=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-O3 -mcpu=cortex-a53 -ftree-vectorize -ffat-lto-objects-Ofast -mcpu=cortex-a53-Ofast -mtune=cortex-a57 -ftree-vectorize -funroll-loops-Ofast -mtune=cortex-a57 -ftree-vectorize -funroll-loops-Ofast -mcpu=thunderx -ftree-vectorize1. (CXX) g++ options: -fomit-frame-pointer -fipa-pta -march=armv8-a+crc -ffast-math -mtune=native -flto -lSDL -lstdc++

C-Ray

Total Time

OpenBenchmarking.orgSeconds, Fewer Is BetterC-Ray 1.1Total TimeA53 mtune/vectorize, post-patchA53 vectorize GCC 7.0.1A53 vectorize, earlier buildA53 vectorize, pre-patchA53 vectorize, updatedA53 vectorize/LTO, pre patchA53, post patchA57 vectorize/unrolled GCC 6.3A57 vectorize/unrolled GCC 7.0.1thunderx/vectorize, pre-patch4080120160200SE +/- 0.12, N = 3SE +/- 0.02, N = 3SE +/- 1.00, N = 3SE +/- 0.69, N = 3SE +/- 0.27, N = 3SE +/- 0.17, N = 3SE +/- 0.14, N = 3SE +/- 1.47, N = 3SE +/- 0.78, N = 3SE +/- 1.37, N = 3186.61151.02162.23187.97161.80184.81186.69149.61154.81149.82-Ofast -mtune=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-mcpu=cortex-a53 -ftree-vectorize -flto -ffat-lto-objects-Ofast -mcpu=cortex-a53-Ofast -mtune=cortex-a57 -ftree-vectorize -funroll-loops-Ofast -mtune=cortex-a57 -ftree-vectorize -funroll-loops-Ofast -mcpu=thunderx -ftree-vectorize1. (CC) gcc options: -lm -lpthread -O3 -fomit-frame-pointer -fipa-pta -march=armv8-a+crc

Primesieve

1e12 Prime Number Generation

OpenBenchmarking.orgSeconds, Fewer Is BetterPrimesieve 5.4.21e12 Prime Number GenerationA53 mtune/vectorize, post-patchA53 vectorize GCC 7.0.1A53 vectorize, earlier buildA53 vectorize, pre-patchA53 vectorize, updatedA53 vectorize/LTO, pre patchA53, post patchA57 vectorize/unrolled GCC 6.3A57 vectorize/unrolled GCC 7.0.1thunderx/vectorize, pre-patch120240360480600SE +/- 6.92, N = 3SE +/- 9.38, N = 3SE +/- 4.17, N = 3SE +/- 3.01, N = 3SE +/- 9.13, N = 4SE +/- 8.42, N = 3SE +/- 9.14, N = 3SE +/- 1.80, N = 3SE +/- 3.16, N = 3SE +/- 2.99, N = 3573.13547.12523.43543.16574.65540.95553.13531.95525.00566.21-Ofast -mtune=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-O3 -mcpu=cortex-a53 -ftree-vectorize -flto -ffat-lto-objects-Ofast -mcpu=cortex-a53-Ofast -mtune=cortex-a57 -ftree-vectorize -funroll-loops-Ofast -mtune=cortex-a57 -ftree-vectorize -funroll-loops-Ofast -mcpu=thunderx -ftree-vectorize1. (CXX) g++ options: -fomit-frame-pointer -fipa-pta -march=armv8-a+crc -fopenmp

Smallpt

Global Illumination Renderer; 100 Samples

OpenBenchmarking.orgSeconds, Fewer Is BetterSmallpt 1.0Global Illumination Renderer; 100 SamplesA53 mtune/vectorize, post-patchA53 vectorize GCC 7.0.1A53 vectorize, earlier buildA53 vectorize, pre-patchA53 vectorize, updatedA53 vectorize/LTO, pre patchA53, post patchA57 vectorize/unrolled GCC 6.3A57 vectorize/unrolled GCC 7.0.1thunderx/vectorize, pre-patch4080120160200SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 2.33, N = 3SE +/- 0.00, N = 3167166166167166168168169168167-Ofast -mtune=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-O3 -mcpu=cortex-a53 -ftree-vectorize -flto -ffat-lto-objects-Ofast -mcpu=cortex-a53-Ofast -mtune=cortex-a57 -ftree-vectorize -funroll-loops-Ofast -mtune=cortex-a57 -ftree-vectorize -funroll-loops-Ofast -mcpu=thunderx -ftree-vectorize1. (CXX) g++ options: -fopenmp -fomit-frame-pointer -fipa-pta -march=armv8-a+crc

Sudokut

Total Time

OpenBenchmarking.orgSeconds, Fewer Is BetterSudokut 0.4Total TimeA53 mtune/vectorize, post-patchA53 vectorize GCC 7.0.1A53 vectorize, earlier buildA53 vectorize, pre-patchA53 vectorize, updatedA53 vectorize/LTO, pre patchA53, post patchA57 vectorize/unrolled GCC 6.3A57 vectorize/unrolled GCC 7.0.1thunderx/vectorize, pre-patch20406080100SE +/- 0.09, N = 3SE +/- 0.13, N = 3SE +/- 0.06, N = 3SE +/- 0.20, N = 3SE +/- 0.20, N = 3SE +/- 0.21, N = 3SE +/- 0.10, N = 3SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.76, N = 3102.17102.97102.59101.95102.72101.75101.88103.00103.04102.75

Tachyon

Total Time

OpenBenchmarking.orgSeconds, Fewer Is BetterTachyon 0.98.9Total TimeA53 mtune/vectorize, post-patchA53 vectorize GCC 7.0.1A53 vectorize, earlier buildA53 vectorize, pre-patchA53 vectorize, updatedA53 vectorize/LTO, pre patchA53, post patchA57 vectorize/unrolled GCC 6.3thunderx/vectorize, pre-patch1632486480SE +/- 0.10, N = 3SE +/- 0.17, N = 3SE +/- 0.22, N = 3SE +/- 0.08, N = 3SE +/- 0.29, N = 3SE +/- 0.11, N = 3SE +/- 0.12, N = 3SE +/- 0.03, N = 3SE +/- 0.06, N = 369.3469.6469.3969.2769.9067.6469.4072.2871.41

OpenSSL

RSA 4096-bit Performance

OpenBenchmarking.orgSigns Per Second, More Is BetterOpenSSL 1.0.1gRSA 4096-bit PerformanceA53 mtune/vectorize, post-patchA53 vectorize GCC 7.0.1A53 vectorize, earlier buildA53 vectorize, pre-patchA53 vectorize, updatedA53 vectorize/LTO, pre patchA53, post patchA57 vectorize/unrolled GCC 6.3A57 vectorize/unrolled GCC 7.0.1thunderx/vectorize, pre-patch510152025SE +/- 0.00, N = 3SE +/- 0.03, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.03, N = 3SE +/- 0.00, N = 321.5021.4721.4021.5021.4021.5021.5021.2021.4721.501. (CC) gcc options: -O3 -fomit-frame-pointer -lssl -lcrypto -ldl

Redis

Test: GET

OpenBenchmarking.orgRequests Per Second, More Is BetterRedis 3.0.1Test: GETA53 mtune/vectorize, post-patchA53 vectorize GCC 7.0.1A53 vectorize, earlier buildA53 vectorize, pre-patchA53 vectorize, updatedA53 vectorize/LTO, pre patchA53, post patchA57 vectorize/unrolled GCC 6.3A57 vectorize/unrolled GCC 7.0.1thunderx/vectorize, pre-patch70K140K210K280K350KSE +/- 1967.34, N = 3SE +/- 2031.32, N = 3SE +/- 419.32, N = 3SE +/- 4662.92, N = 6SE +/- 2017.17, N = 3SE +/- 2239.53, N = 3SE +/- 1052.91, N = 3SE +/- 3267.58, N = 3SE +/- 649.43, N = 3SE +/- 2784.59, N = 3313438.91276298.44283742.83310344.73277268.23311785.02309030.64275458.70276169.86318926.02-Ofast -mtune=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-Ofast -mcpu=cortex-a53 -ftree-vectorize-O3 -mcpu=cortex-a53 -ftree-vectorize -flto -ffat-lto-objects-Ofast -mcpu=cortex-a53-Ofast -mtune=cortex-a57 -ftree-vectorize -funroll-loops-Ofast -mtune=cortex-a57 -ftree-vectorize -funroll-loops-Ofast -mcpu=thunderx -ftree-vectorize1. (CC) gcc options: -ggdb -rdynamic -lm -pthread -ldl -O2 -fomit-frame-pointer -fipa-pta -march=armv8-a+crc

Phoronix Test Suite v10.8.4