Amazon EC2 A1 ARM Linux Benchmarks

KVM testing on Ubuntu 18.04 via the Phoronix Test Suite.

HTML result view exported from: https://openbenchmarking.org/result/1811283-SK-1811286SK95&grs&export=pdf&rdt&rro.

Amazon EC2 A1 ARM Linux BenchmarksProcessorMotherboardChipsetMemoryDiskNetworkOSKernelCompilerFile-SystemSystem LayerARMv8 a1.largeARMv8 a1.xlargeARMv8 a1.2xlargeARMv8 a1.4xlargeAMD m5a.xlargeAMD m5a.4xlargeAMD m5a.largeAMD m5a.2xlargeARMv8 Cortex-A72 (2 Cores)Amazon EC2 a1.large (1.0 BIOS)Amazon Device 02004096MB52GB Amazon Elastic Block StoreAmazon ElasticUbuntu 18.044.15.0-1028-aws (aarch64)GCC 7.3.0ext4ARMv8 Cortex-A72 (4 Cores)Amazon EC2 a1.xlarge (1.0 BIOS)8192MBARMv8 Cortex-A72 (8 Cores)Amazon EC2 a1.2xlarge (1.0 BIOS)16384MBARMv8 Cortex-A72 (16 Cores)Amazon EC2 a1.4xlarge (1.0 BIOS)31744MBAMD EPYC 7571 (2 Cores / 4 Threads)Amazon EC2 m5a.xlarge (1.0 BIOS)Intel 440FX 82441FX PMC16384MB4.15.0-1021-aws (x86_64)KVMAMD EPYC 7571 (8 Cores / 16 Threads)Amazon EC2 m5a.4xlarge (1.0 BIOS)63488MBAMD EPYC 7571 (1 Core / 2 Threads)Amazon EC2 m5a.large (1.0 BIOS)8192MBAMD EPYC 7571 (4 Cores / 8 Threads)Amazon EC2 m5a.2xlarge (1.0 BIOS)31744MBOpenBenchmarking.orgCompiler Details- ARMv8 a1.large: --build=aarch64-linux-gnu --disable-libquadmath --disable-libquadmath-support --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-fix-cortex-a53-843419 --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++ --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-plugin --enable-shared --enable-threads=posix --host=aarch64-linux-gnu --program-prefix=aarch64-linux-gnu- --target=aarch64-linux-gnu --with-default-libstdcxx-abi=new --with-gcc-major-version-only -v - ARMv8 a1.xlarge: --build=aarch64-linux-gnu --disable-libquadmath --disable-libquadmath-support --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-fix-cortex-a53-843419 --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++ --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-plugin --enable-shared --enable-threads=posix --host=aarch64-linux-gnu --program-prefix=aarch64-linux-gnu- --target=aarch64-linux-gnu --with-default-libstdcxx-abi=new --with-gcc-major-version-only -v - ARMv8 a1.2xlarge: --build=aarch64-linux-gnu --disable-libquadmath --disable-libquadmath-support --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-fix-cortex-a53-843419 --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++ --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-plugin --enable-shared --enable-threads=posix --host=aarch64-linux-gnu --program-prefix=aarch64-linux-gnu- --target=aarch64-linux-gnu --with-default-libstdcxx-abi=new --with-gcc-major-version-only -v - ARMv8 a1.4xlarge: --build=aarch64-linux-gnu --disable-libquadmath --disable-libquadmath-support --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-fix-cortex-a53-843419 --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++ --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-plugin --enable-shared --enable-threads=posix --host=aarch64-linux-gnu --program-prefix=aarch64-linux-gnu- --target=aarch64-linux-gnu --with-default-libstdcxx-abi=new --with-gcc-major-version-only -v - AMD m5a.xlarge: --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++ --enable-libmpx --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none --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 --with-tune=generic --without-cuda-driver -v - AMD m5a.4xlarge: --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++ --enable-libmpx --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none --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 --with-tune=generic --without-cuda-driver -v - AMD m5a.large: --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++ --enable-libmpx --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none --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 --with-tune=generic --without-cuda-driver -v - AMD m5a.2xlarge: --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++ --enable-libmpx --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none --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 --with-tune=generic --without-cuda-driver -v Python Details- Python 2.7.15rc1 + Python 3.6.7Security Details- AMD m5a.xlarge, AMD m5a.4xlarge, AMD m5a.large, AMD m5a.2xlarge: __user pointer sanitization + Full generic retpoline

Amazon EC2 A1 ARM Linux Benchmarksx264: H.264 Video Encodingrodinia: OpenMP LavaMDm-queens: Time To Solvego-benchmark: garbagego-benchmark: jsoncompress-7zip: Compress Speed Testrodinia: OpenMP CFD Solvercompress-zstd: Compressing ubuntu-16.04.3-server-i386.img, Compression Level 19rust-mandel: Time To Complete Serial/Parallel Mandelbrotrust-prime: Prime Number Test To 200,000,000php: Zend micro_benchpybench: Total For Average Test Timesapache: Static Web Page Servingphp: Zend benchperl-benchmark: Interpreterphpbench: PHP Benchmark Suiteperl-benchmark: Pod2htmlgo-benchmark: buildencode-mp3: WAV To MP3ARMv8 a1.largeARMv8 a1.xlargeARMv8 a1.2xlargeARMv8 a1.4xlargeAMD m5a.xlargeAMD m5a.4xlargeAMD m5a.largeAMD m5a.2xlarge5.38107488216850985788421923868181171243128.308.94426655421.810.002430511688550.425048153959161198383.118.76541441871506339753346686611110815264.228.88424474841.810.002436181684670.435032333967382596683.1216.622722214727623203203181358961.8962.0410632.089.04428971111.820.002383171674650.430430234003113767183.2632.431381102599315105408412523040.9241.8583.9816.069.09427272311.790.002404821673440.429817224134090511583.0316.20443392702381233246787909190.4386.05138.01231.783.191664138710.730.001058543693620.206644512075875822647.0250.901311142164682103406552855734.6033.2774.4469.333.111655102740.740.001316573664110.205557482109779824247.059.5188678613241728669863504711169.13161.43228.74463.023.14165792250.740.001058983656650.212059712153465731547.9226.542632334112557200445691470258.6054.93100.95139.883.261662105340.840.001152483425530.205777412080208664948.09OpenBenchmarking.org

Apache Benchmark

Performance / Cost - Static Web Page Serving

OpenBenchmarking.orgRequests Per Second Per Dollar, More Is BetterApache Benchmark 2.4.29Performance / Cost - Static Web Page ServingAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large300K600K900K1200K1500K351143.67922537.00171233.331387149.00144625.40237034.00748416.00554229.001. AMD m5a.2xlarge: $0.344 reported cost per hour, test consumed 5 Minutes, 5 Seconds: cost approximately 0.03 dollar.2. AMD m5a.large: $0.086 reported cost per hour, test consumed 5 Minutes, 45 Seconds: cost approximately 0.01 dollar.3. AMD m5a.4xlarge: $0.688 reported cost per hour, test consumed 5 Minutes, 12 Seconds: cost approximately 0.06 dollar. 4. AMD m5a.xlarge: $0.172 reported cost per hour, test consumed 3 Minutes, 57 Seconds: cost approximately 0.01 dollar.5. ARMv8 a1.4xlarge: $0.4080 reported cost per hour, test consumed 7 Minutes, 17 Seconds: cost approximately 0.05 dollar. 6. ARMv8 a1.2xlarge: $0.2040 reported cost per hour, test consumed 7 Minutes, 25 Seconds: cost approximately 0.03 dollar. 7. ARMv8 a1.xlarge: $0.1020 reported cost per hour, test consumed 7 Minutes, 12 Seconds: cost approximately 0.01 dollar. 8. ARMv8 a1.large: $0.0510 reported cost per hour, test consumed 9 Minutes, 34 Seconds: cost approximately 0.01 dollar.

x264

H.264 Video Encoding

OpenBenchmarking.orgFrames Per Second, More Is Betterx264 2018-09-25H.264 Video EncodingAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large1122334455SE +/- 0.07, N = 3SE +/- 0.01, N = 3SE +/- 0.17, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.03, N = 326.549.5150.9016.2032.4316.628.765.38-m64 -O3 -ffast-math -std=gnu99 -fPIC -fomit-frame-pointer -fno-tree-vectorize-m64 -O3 -ffast-math -std=gnu99 -fPIC -fomit-frame-pointer -fno-tree-vectorize-m64 -O3 -ffast-math -std=gnu99 -fPIC -fomit-frame-pointer -fno-tree-vectorize-m64 -O3 -ffast-math -std=gnu99 -fPIC -fomit-frame-pointer -fno-tree-vectorize1. (CC) gcc options: -ldl -lm -lpthread

Rust Prime Benchmark

Performance / Cost - Prime Number Test To 200,000,000

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterRust Prime BenchmarkPerformance / Cost - Prime Number Test To 200,000,000AMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large481216205.6013.892.776.950.160.320.641.281. AMD m5a.2xlarge: $0.344 reported cost per hour, test consumed 7 Minutes, 3 Seconds: cost approximately 0.04 dollar.2. AMD m5a.large: $0.086 reported cost per hour, test consumed 23 Minutes, 13 Seconds: cost approximately 0.03 dollar.3. AMD m5a.4xlarge: $0.688 reported cost per hour, test consumed 3 Minutes, 32 Seconds: cost approximately 0.04 dollar. 4. AMD m5a.xlarge: $0.172 reported cost per hour, test consumed 11 Minutes, 39 Seconds: cost approximately 0.03 dollar. 5. ARMv8 a1.4xlarge: $0.4080 reported cost per hour, test consumed 52 Seconds: cost approximately 0.01 dollar.6. ARMv8 a1.2xlarge: $0.2040 reported cost per hour, test consumed 1 Minute, 40 Seconds: cost approximately 0.01 dollar. 7. ARMv8 a1.xlarge: $0.1020 reported cost per hour, test consumed 3 Minutes, 17 Seconds: cost approximately 0.01 dollar. 8. ARMv8 a1.large: $0.0510 reported cost per hour, test consumed 6 Minutes, 29 Seconds: cost approximately 0.01 dollar.

Rodinia

Test: OpenMP LavaMD

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 2.4Test: OpenMP LavaMDAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large2004006008001000SE +/- 0.24, N = 3SE +/- 0.79, N = 3SE +/- 0.14, N = 3SE +/- 0.52, N = 3SE +/- 0.09, N = 3SE +/- 0.07, N = 3SE +/- 0.23, N = 3SE +/- 1.41, N = 326388613144313827254110741. (CXX) g++ options: -O2 -lOpenCL

m-queens

Time To Solve

OpenBenchmarking.orgSeconds, Fewer Is Betterm-queens 1.2Time To SolveAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large2004006008001000SE +/- 0.09, N = 3SE +/- 0.24, N = 3SE +/- 0.01, N = 3SE +/- 0.12, N = 3SE +/- 0.01, N = 3SE +/- 0.07, N = 3SE +/- 0.06, N = 3SE +/- 0.10, N = 32337861143921102214418821. (CXX) g++ options: -fopenmp -O2 -march=native

Rodinia

Performance / Cost - Test: OpenMP LavaMD

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterRodinia 2.4Performance / Cost - Test: OpenMP LavaMDAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large122436486021.0053.1410.4726.616.8913.6027.0653.721. AMD m5a.2xlarge: $0.344 reported cost per hour, test consumed 13 Minutes, 12 Seconds: cost approximately 0.08 dollar. 2. AMD m5a.large: $0.086 reported cost per hour, test consumed 44 Minutes, 21 Seconds: cost approximately 0.06 dollar.3. AMD m5a.4xlarge: $0.688 reported cost per hour, test consumed 6 Minutes, 37 Seconds: cost approximately 0.08 dollar. 4. AMD m5a.xlarge: $0.172 reported cost per hour, test consumed 22 Minutes, 14 Seconds: cost approximately 0.06 dollar. 5. ARMv8 a1.4xlarge: $0.4080 reported cost per hour, test consumed 6 Minutes, 57 Seconds: cost approximately 0.05 dollar. 6. ARMv8 a1.2xlarge: $0.2040 reported cost per hour, test consumed 13 Minutes, 40 Seconds: cost approximately 0.05 dollar. 7. ARMv8 a1.xlarge: $0.1020 reported cost per hour, test consumed 27 Minutes, 14 Seconds: cost approximately 0.05 dollar. 8. ARMv8 a1.large: $0.0510 reported cost per hour, test consumed 53 Minutes, 53 Seconds: cost approximately 0.05 dollar.

Go Benchmarks

Test: garbage

OpenBenchmarking.orgNanoseconds/Operation, Fewer Is BetterGo BenchmarksTest: garbageAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large4M8M12M16M20MSE +/- 9450.61, N = 3SE +/- 30608.39, N = 3SE +/- 5851.87, N = 3SE +/- 21715.41, N = 3SE +/- 1337.70, N = 3SE +/- 1743.82, N = 3SE +/- 7895.81, N = 3SE +/- 48615.61, N = 34112557132417282164682702381225993154727623871506316850985

PyBench

Performance / Cost - Total For Average Test Times

OpenBenchmarking.orgMilliseconds x Dollar, Fewer Is BetterPyBench 2018-02-16Performance / Cost - Total For Average Test TimesAMD m5a.2xlargeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlarge30609012015016.6233.1016.64128.1685.7842.441. AMD m5a.2xlarge: $0.344 reported cost per hour, test consumed 2 Minutes, 7 Seconds: cost approximately 0.01 dollar.2. AMD m5a.4xlarge: $0.688 reported cost per hour, test consumed 1 Minute, 57 Seconds: cost approximately 0.02 dollar.3. AMD m5a.xlarge: $0.172 reported cost per hour, test consumed 1 Minute, 58 Seconds: cost approximately 0.01 dollar.4. ARMv8 a1.4xlarge: $0.4080 reported cost per hour, test consumed 4 Minutes, 48 Seconds: cost approximately 0.03 dollar. 5. ARMv8 a1.2xlarge: $0.2040 reported cost per hour, test consumed 4 Minutes, 49 Seconds: cost approximately 0.02 dollar. 6. ARMv8 a1.xlarge: $0.1020 reported cost per hour, test consumed 4 Minutes, 47 Seconds: cost approximately 0.01 dollar.

Go Benchmarks

Test: json

OpenBenchmarking.orgNanoseconds/Operation, Fewer Is BetterGo BenchmarksTest: jsonAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large20M40M60M80M100MSE +/- 92676.78, N = 3SE +/- 162025.72, N = 3SE +/- 33427.39, N = 3SE +/- 95282.30, N = 3SE +/- 42977.01, N = 3SE +/- 125406.02, N = 3SE +/- 148481.92, N = 3SE +/- 387947.87, N = 32004456966986350103406553324678710540841203203183975334678842192

7-Zip Compression

Compress Speed Test

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 16.02Compress Speed TestAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large6K12K18K24K30KSE +/- 80.50, N = 3SE +/- 6.24, N = 3SE +/- 21.61, N = 3SE +/- 41.25, N = 3SE +/- 123.49, N = 3SE +/- 39.96, N = 3SE +/- 18.00, N = 3SE +/- 6.57, N = 31470247112855790912523013589686638681. (CXX) g++ options: -pipe -lpthread

x264

Performance / Cost - H.264 Video Encoding

OpenBenchmarking.orgFrames Per Second Per Dollar, More Is Betterx264 2018-09-25Performance / Cost - H.264 Video EncodingAMD m5a.2xlargeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlarge110022003300440055002654.005090.001620.003243.001662.00876.001. AMD m5a.2xlarge: $0.344 reported cost per hour, test consumed 1 Minute, 12 Seconds: cost approximately 0.01 dollar.2. AMD m5a.4xlarge: $0.688 reported cost per hour, test consumed 39 Seconds: cost approximately 0.01 dollar.3. AMD m5a.xlarge: $0.172 reported cost per hour, test consumed 1 Minute, 55 Seconds: cost approximately 0.01 dollar.4. ARMv8 a1.4xlarge: $0.4080 reported cost per hour, test consumed 1 Minute: cost approximately 0.01 dollar.5. ARMv8 a1.2xlarge: $0.2040 reported cost per hour, test consumed 1 Minute, 53 Seconds: cost approximately 0.01 dollar. 6. ARMv8 a1.xlarge: $0.1020 reported cost per hour, test consumed 3 Minutes, 44 Seconds: cost approximately 0.01 dollar.

LAME MP3 Encoding

Performance / Cost - WAV To MP3

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterLAME MP3 Encoding 3.100Performance / Cost - WAV To MP3AMD m5a.2xlargeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlarge0.56031.12061.68092.24122.80150.481.410.472.490.830.831. AMD m5a.2xlarge: $0.344 reported cost per hour, test consumed 2 Minutes, 28 Seconds: cost approximately 0.01 dollar. 2. AMD m5a.4xlarge: $0.688 reported cost per hour, test consumed 2 Minutes, 25 Seconds: cost approximately 0.03 dollar. 3. AMD m5a.xlarge: $0.172 reported cost per hour, test consumed 2 Minutes, 25 Seconds: cost approximately 0.01 dollar.4. ARMv8 a1.4xlarge: $0.4080 reported cost per hour, test consumed 4 Minutes, 13 Seconds: cost approximately 0.03 dollar. 5. ARMv8 a1.2xlarge: $0.2040 reported cost per hour, test consumed 4 Minutes, 14 Seconds: cost approximately 0.01 dollar. 6. ARMv8 a1.xlarge: $0.1020 reported cost per hour, test consumed 4 Minutes, 14 Seconds: cost approximately 0.01 dollar.

Rodinia

Test: OpenMP CFD Solver

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 2.4Test: OpenMP CFD SolverAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large4080120160200SE +/- 0.04, N = 3SE +/- 0.76, N = 3SE +/- 0.23, N = 3SE +/- 0.13, N = 3SE +/- 0.09, N = 3SE +/- 0.10, N = 3SE +/- 0.24, N = 3SE +/- 0.37, N = 358.60169.1334.6090.4340.9261.89111.00181.001. (CXX) g++ options: -O2 -lOpenCL

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 19AMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large4080120160200SE +/- 0.17, N = 3SE +/- 1.27, N = 3SE +/- 0.02, N = 3SE +/- 0.13, N = 3SE +/- 0.04, N = 3SE +/- 0.09, N = 3SE +/- 2.34, N = 3SE +/- 2.66, N = 354.93161.4333.2786.0541.8562.04108.00171.00-lz1. (CC) gcc options: -O3 -pthread

Rodinia

Performance / Cost - Test: OpenMP CFD Solver

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterRodinia 2.4Performance / Cost - Test: OpenMP CFD SolverAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large0.40730.81461.22191.62922.03651.171.690.690.900.410.621.111.811. AMD m5a.2xlarge: $0.344 reported cost per hour, test consumed 3 Minutes: cost approximately 0.02 dollar.2. AMD m5a.large: $0.086 reported cost per hour, test consumed 8 Minutes, 31 Seconds: cost approximately 0.01 dollar.3. AMD m5a.4xlarge: $0.688 reported cost per hour, test consumed 1 Minute, 48 Seconds: cost approximately 0.02 dollar.4. AMD m5a.xlarge: $0.172 reported cost per hour, test consumed 4 Minutes, 36 Seconds: cost approximately 0.01 dollar.5. ARMv8 a1.4xlarge: $0.4080 reported cost per hour, test consumed 2 Minutes, 6 Seconds: cost approximately 0.01 dollar. 6. ARMv8 a1.2xlarge: $0.2040 reported cost per hour, test consumed 3 Minutes, 10 Seconds: cost approximately 0.01 dollar. 7. ARMv8 a1.xlarge: $0.1020 reported cost per hour, test consumed 5 Minutes, 38 Seconds: cost approximately 0.01 dollar. 8. ARMv8 a1.large: $0.0510 reported cost per hour, test consumed 9 Minutes, 8 Seconds: cost approximately 0.01 dollar.

PHPBench

Performance / Cost - PHP Benchmark Suite

OpenBenchmarking.orgScore Per Dollar, More Is BetterPHPBench 0.8.1Performance / Cost - PHP Benchmark SuiteAMD m5a.2xlargeAMD m5a.4xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlarge7M14M21M28M35M34255300.0018320550.008367200.0016746500.0016846700.001. AMD m5a.2xlarge: $0.344 reported cost per hour, test consumed 1 Minute, 32 Seconds: cost approximately 0.01 dollar.2. AMD m5a.4xlarge: $0.688 reported cost per hour, test consumed 1 Minute, 26 Seconds: cost approximately 0.02 dollar.3. ARMv8 a1.4xlarge: $0.4080 reported cost per hour, test consumed 3 Minutes, 4 Seconds: cost approximately 0.02 dollar. 4. ARMv8 a1.2xlarge: $0.2040 reported cost per hour, test consumed 3 Minutes, 3 Seconds: cost approximately 0.01 dollar. 5. ARMv8 a1.xlarge: $0.1020 reported cost per hour, test consumed 3 Minutes, 9 Seconds: cost approximately 0.01 dollar.

Zstd Compression

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

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterZstd Compression 1.3.4Performance / Cost - Compressing ubuntu-16.04.3-server-i386.img, Compression Level 19AMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large0.38480.76961.15441.53921.9241.101.610.670.860.420.621.081.711. AMD m5a.2xlarge: $0.344 reported cost per hour, test consumed 2 Minutes, 50 Seconds: cost approximately 0.02 dollar. 2. AMD m5a.large: $0.086 reported cost per hour, test consumed 8 Minutes, 10 Seconds: cost approximately 0.01 dollar.3. AMD m5a.4xlarge: $0.688 reported cost per hour, test consumed 1 Minute, 44 Seconds: cost approximately 0.02 dollar.4. AMD m5a.xlarge: $0.172 reported cost per hour, test consumed 4 Minutes, 23 Seconds: cost approximately 0.01 dollar.5. ARMv8 a1.4xlarge: $0.4080 reported cost per hour, test consumed 2 Minutes, 10 Seconds: cost approximately 0.01 dollar. 6. ARMv8 a1.2xlarge: $0.2040 reported cost per hour, test consumed 3 Minutes, 11 Seconds: cost approximately 0.01 dollar. 7. ARMv8 a1.xlarge: $0.1020 reported cost per hour, test consumed 5 Minutes, 31 Seconds: cost approximately 0.01 dollar. 8. ARMv8 a1.large: $0.0510 reported cost per hour, test consumed 8 Minutes, 39 Seconds: cost approximately 0.01 dollar.

Rust Mandelbrot

Time To Complete Serial/Parallel Mandelbrot

OpenBenchmarking.orgSeconds, Fewer Is BetterRust MandelbrotTime To Complete Serial/Parallel MandelbrotAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large50100150200250SE +/- 0.12, N = 3SE +/- 0.67, N = 3SE +/- 0.23, N = 3SE +/- 0.06, N = 3SE +/- 0.28, N = 3SE +/- 0.00, N = 3SE +/- 0.07, N = 3SE +/- 0.03, N = 3100.95228.7474.44138.0183.98106.00152.00243.00-m64-m64-m64-m641. (CC) gcc options: -pie -nodefaultlibs

Rust Mandelbrot

Performance / Cost - Time To Complete Serial/Parallel Mandelbrot

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterRust MandelbrotPerformance / Cost - Time To Complete Serial/Parallel MandelbrotAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large1.02832.05663.08494.11325.14153.034.572.982.762.522.131.522.431. AMD m5a.2xlarge: $0.344 reported cost per hour, test consumed 5 Minutes, 6 Seconds: cost approximately 0.03 dollar.2. AMD m5a.large: $0.086 reported cost per hour, test consumed 11 Minutes, 30 Seconds: cost approximately 0.02 dollar.3. AMD m5a.4xlarge: $0.688 reported cost per hour, test consumed 3 Minutes, 47 Seconds: cost approximately 0.04 dollar. 4. AMD m5a.xlarge: $0.172 reported cost per hour, test consumed 6 Minutes, 58 Seconds: cost approximately 0.02 dollar.5. ARMv8 a1.4xlarge: $0.4080 reported cost per hour, test consumed 4 Minutes, 16 Seconds: cost approximately 0.03 dollar. 6. ARMv8 a1.2xlarge: $0.2040 reported cost per hour, test consumed 5 Minutes, 23 Seconds: cost approximately 0.02 dollar. 7. ARMv8 a1.xlarge: $0.1020 reported cost per hour, test consumed 7 Minutes, 41 Seconds: cost approximately 0.01 dollar. 8. ARMv8 a1.large: $0.0510 reported cost per hour, test consumed 12 Minutes, 15 Seconds: cost approximately 0.01 dollar.

Rust Prime Benchmark

Prime Number Test To 200,000,000

OpenBenchmarking.orgSeconds, Fewer Is BetterRust Prime BenchmarkPrime Number Test To 200,000,000AMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large100200300400500SE +/- 0.09, N = 3SE +/- 1.35, N = 3SE +/- 0.13, N = 3SE +/- 0.08, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.08, N = 3SE +/- 0.01, N = 3139.88463.0269.33231.7816.0632.0864.22128.30-m64-m64-m64-m641. (CC) gcc options: -pie -nodefaultlibs

Cost To Run Tests

Cost / Price Per Hour

OpenBenchmarking.orgCost / Price Per Hour, Fewer Is BetterCost To Run TestsCost / Price Per HourAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large0.1260.2520.3780.5040.630.400.230.560.270.400.260.190.251. AMD m5a.2xlarge: $0.344 reported cost per hour, running tests consumed 1 Hour, 9 Minutes, 22 Seconds: cost approximately 0.4 dollar. 2. AMD m5a.large: $0.086 reported cost per hour, running tests consumed 2 Hours, 43 Minutes: cost approximately 0.23 dollar. 3. AMD m5a.4xlarge: $0.688 reported cost per hour, running tests consumed 49 Minutes, 4 Seconds: cost approximately 0.56 dollar. 4. AMD m5a.xlarge: $0.172 reported cost per hour, running tests consumed 1 Hour, 35 Minutes, 29 Seconds: cost approximately 0.27 dollar. 5. ARMv8 a1.4xlarge: $0.4080 reported cost per hour, running tests consumed 59 Minutes, 8 Seconds: cost approximately 0.4 dollar. 6. ARMv8 a1.2xlarge: $0.2040 reported cost per hour, running tests consumed 1 Hour, 15 Minutes, 53 Seconds: cost approximately 0.26 dollar. 7. ARMv8 a1.xlarge: $0.1020 reported cost per hour, running tests consumed 1 Hour, 50 Minutes, 55 Seconds: cost approximately 0.19 dollar. 8. ARMv8 a1.large: $0.0510 reported cost per hour, running tests consumed 4 Hours, 50 Minutes, 30 Seconds: cost approximately 0.25 dollar.

PHP Micro Benchmarks

Test: Zend micro_bench

OpenBenchmarking.orgSeconds, Fewer Is BetterPHP Micro BenchmarksTest: Zend micro_benchAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large3691215SE +/- 0.04, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.21, N = 3SE +/- 0.31, N = 3SE +/- 0.28, N = 3SE +/- 0.33, N = 33.263.143.113.199.099.048.888.94

7-Zip Compression

Performance / Cost - Compress Speed Test

OpenBenchmarking.orgMIPS Per Dollar, More Is Better7-Zip Compression 16.02Performance / Cost - Compress Speed TestAMD m5a.2xlargeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlarge500K1000K1500K2000K2500K1470200.00951900.00909100.002523000.001358900.001. AMD m5a.2xlarge: $0.344 reported cost per hour, test consumed 2 Minutes, 18 Seconds: cost approximately 0.01 dollar. 2. AMD m5a.4xlarge: $0.688 reported cost per hour, test consumed 2 Minutes, 24 Seconds: cost approximately 0.03 dollar. 3. AMD m5a.xlarge: $0.172 reported cost per hour, test consumed 1 Minute, 53 Seconds: cost approximately 0.01 dollar.4. ARMv8 a1.4xlarge: $0.4080 reported cost per hour, test consumed 2 Minutes, 10 Seconds: cost approximately 0.01 dollar. 5. ARMv8 a1.2xlarge: $0.2040 reported cost per hour, test consumed 1 Minute, 58 Seconds: cost approximately 0.01 dollar.

PyBench

Total For Average Test Times

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyBench 2018-02-16Total For Average Test TimesAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large9001800270036004500SE +/- 0.67, N = 3SE +/- 8.50, N = 2SE +/- 3.48, N = 3SE +/- 4.37, N = 3SE +/- 38.48, N = 3SE +/- 37.62, N = 3SE +/- 5.33, N = 3SE +/- 6.33, N = 316621657165516644272428942444266

Apache Benchmark

Static Web Page Serving

OpenBenchmarking.orgRequests Per Second, More Is BetterApache Benchmark 2.4.29Static Web Page ServingAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large3K6K9K12K15KSE +/- 24.40, N = 3SE +/- 59.99, N = 3SE +/- 133.49, N = 3SE +/- 575.09, N = 3SE +/- 5.29, N = 3SE +/- 38.92, N = 3SE +/- 98.43, N = 3SE +/- 17.12, N = 3105349225102741387172317111748455421. (CC) gcc options: -shared -fPIC -O2 -pthread

PHP Micro Benchmarks

Test: Zend bench

OpenBenchmarking.orgSeconds, Fewer Is BetterPHP Micro BenchmarksTest: Zend benchAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large0.40950.8191.22851.6382.0475SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.840.740.740.731.791.821.811.81

Perl Benchmarks

Test: Interpreter

OpenBenchmarking.orgSeconds, Fewer Is BetterPerl BenchmarksTest: InterpreterAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large0.00050.0010.00150.0020.0025SE +/- 0.00000292, N = 3SE +/- 0.00000220, N = 3SE +/- 0.00000154, N = 3SE +/- 0.00000060, N = 3SE +/- 0.00000487, N = 3SE +/- 0.00000631, N = 3SE +/- 0.00000575, N = 3SE +/- 0.00000400, N = 30.001152480.001058980.001316570.001058540.002404820.002383170.002436180.00243051

PHPBench

PHP Benchmark Suite

OpenBenchmarking.orgScore, More Is BetterPHPBench 0.8.1PHP Benchmark SuiteAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large80K160K240K320K400KSE +/- 6565.05, N = 3SE +/- 1991.74, N = 3SE +/- 2728.14, N = 3SE +/- 656.89, N = 3SE +/- 919.84, N = 3SE +/- 725.88, N = 3SE +/- 554.92, N = 3SE +/- 131.93, N = 3342553365665366411369362167344167465168467168855

Perl Benchmarks

Test: Pod2html

OpenBenchmarking.orgSeconds, Fewer Is BetterPerl BenchmarksTest: Pod2htmlAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large0.09790.19580.29370.39160.4895SE +/- 0.00084270, N = 3SE +/- 0.00618143, N = 3SE +/- 0.00088366, N = 3SE +/- 0.00030351, N = 3SE +/- 0.00165742, N = 3SE +/- 0.00449263, N = 3SE +/- 0.00200775, N = 3SE +/- 0.00099623, N = 30.205777410.212059710.205557480.206644510.429817220.430430230.435032330.42504815

Perl Benchmarks

Performance / Cost - Test: Pod2html

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterPerl BenchmarksPerformance / Cost - Test: Pod2htmlAMD m5a.4xlargeARMv8 a1.4xlargeARMv8 a1.2xlarge0.00450.0090.01350.0180.02250.010.020.011. AMD m5a.4xlarge: $0.688 reported cost per hour, test consumed 4 Minutes, 13 Seconds: cost approximately 0.05 dollar. 2. ARMv8 a1.4xlarge: $0.4080 reported cost per hour, test consumed 5 Minutes, 40 Seconds: cost approximately 0.04 dollar. 3. ARMv8 a1.2xlarge: $0.2040 reported cost per hour, test consumed 5 Minutes, 40 Seconds: cost approximately 0.02 dollar.

m-queens

Performance / Cost - Time To Solve

OpenBenchmarking.orgSeconds x Dollar, Fewer Is Betterm-queens 1.2Performance / Cost - Time To SolveAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large112233445516.2947.198.0023.544.428.8317.6535.291. AMD m5a.2xlarge: $0.344 reported cost per hour, test consumed 11 Minutes, 43 Seconds: cost approximately 0.07 dollar. 2. AMD m5a.large: $0.086 reported cost per hour, test consumed 39 Minutes, 24 Seconds: cost approximately 0.06 dollar.3. AMD m5a.4xlarge: $0.688 reported cost per hour, test consumed 5 Minutes, 47 Seconds: cost approximately 0.07 dollar. 4. AMD m5a.xlarge: $0.172 reported cost per hour, test consumed 19 Minutes, 41 Seconds: cost approximately 0.06 dollar. 5. ARMv8 a1.4xlarge: $0.4080 reported cost per hour, test consumed 5 Minutes, 36 Seconds: cost approximately 0.04 dollar. 6. ARMv8 a1.2xlarge: $0.2040 reported cost per hour, test consumed 11 Minutes, 6 Seconds: cost approximately 0.04 dollar. 7. ARMv8 a1.xlarge: $0.1020 reported cost per hour, test consumed 22 Minutes, 8 Seconds: cost approximately 0.04 dollar. 8. ARMv8 a1.large: $0.0510 reported cost per hour, test consumed 44 Minutes, 12 Seconds: cost approximately 0.04 dollar.

Go Benchmarks

Test: build

OpenBenchmarking.orgNanoseconds/Operation, Fewer Is BetterGo BenchmarksTest: buildAMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large9000M18000M27000M36000M45000MSE +/- 94610444.88, N = 3SE +/- 44282080.00, N = 3SE +/- 29191926.79, N = 3SE +/- 83856991.20, N = 3SE +/- 137165030.87, N = 3SE +/- 87166774.33, N = 3SE +/- 28219638.51, N = 3SE +/- 314217785.70, N = 32080208664921534657315210977982422075875822641340905115400311376713967382596639591611983

Go Benchmarks

Performance / Cost - Test: build

OpenBenchmarking.orgNanoseconds/Operation x Dollar, Fewer Is BetterGo BenchmarksPerformance / Cost - Test: buildAMD m5a.2xlargeAMD m5a.4xlargeARMv8 a1.4xlargeARMv8 a1.2xlarge90M180M270M360M450M208020866.49210977982.42413409051.15400311376.711. AMD m5a.2xlarge: $0.344 reported cost per hour, test consumed 1 Minute, 6 Seconds: cost approximately 0.01 dollar.2. AMD m5a.4xlarge: $0.688 reported cost per hour, test consumed 1 Minute, 7 Seconds: cost approximately 0.01 dollar.3. ARMv8 a1.4xlarge: $0.4080 reported cost per hour, test consumed 2 Minutes, 8 Seconds: cost approximately 0.01 dollar. 4. ARMv8 a1.2xlarge: $0.2040 reported cost per hour, test consumed 2 Minutes, 4 Seconds: cost approximately 0.01 dollar.

LAME MP3 Encoding

WAV To MP3

OpenBenchmarking.orgSeconds, Fewer Is BetterLAME MP3 Encoding 3.100WAV To MP3AMD m5a.2xlargeAMD m5a.largeAMD m5a.4xlargeAMD m5a.xlargeARMv8 a1.4xlargeARMv8 a1.2xlargeARMv8 a1.xlargeARMv8 a1.large20406080100SE +/- 1.27, N = 3SE +/- 0.71, N = 3SE +/- 0.24, N = 3SE +/- 0.02, N = 3SE +/- 0.06, N = 3SE +/- 0.07, N = 3SE +/- 0.04, N = 3SE +/- 0.09, N = 348.0947.9247.0547.0283.0383.2683.1283.111. (CC) gcc options: -lm

Go Benchmarks

Performance / Cost - Test: garbage

OpenBenchmarking.orgNanoseconds/Operation x Dollar, Fewer Is BetterGo BenchmarksPerformance / Cost - Test: garbageAMD m5a.4xlargeARMv8 a1.4xlarge6K12K18K24K30K21646.8225993.151. AMD m5a.4xlarge: $0.688 reported cost per hour, test consumed 42 Seconds: cost approximately 0.01 dollar.2. ARMv8 a1.4xlarge: $0.4080 reported cost per hour, test consumed 51 Seconds: cost approximately 0.01 dollar.

Go Benchmarks

Performance / Cost - Test: json

OpenBenchmarking.orgNanoseconds/Operation x Dollar, Fewer Is BetterGo BenchmarksPerformance / Cost - Test: jsonAMD m5a.4xlarge20K40K60K80K100K103406.551. $0.688 reported cost per hour, test consumed 42 Seconds: cost approximately 0.01 dollar.

Phoronix Test Suite v10.8.5