Amlogic kernel 4.11 vs 3.14 on S905

Memory perfomance comparison on Amlogic S905 (mainline kernel)

HTML result view exported from: https://openbenchmarking.org/result/1704098-RI-1702269RI99.

Amlogic kernel 4.11 vs 3.14 on S905ProcessorMotherboardMemoryDiskGraphicsOSKernelCompilerFile-SystemScreen ResolutionKernel 3.14Kernel 4.11Unknown @ 1.54GHz (4 Cores)Amlogic2048MB16GB NCard + 32GB 00000OSDUbuntu 16.043.14.79-vegas95 (aarch64)GCC 5.4.0 20160609 + Clang 3.8.0-2ubuntu4 + LLVM 3.8.0ext41280x1440Tronsmart Vega S95 Meta16GB SD + 16GB NCard4.11.0-amlogics905x-g2a8c81c-dirty (aarch64)GCC 5.4.0 20160609OpenBenchmarking.orgCompiler Details- --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-gtk-cairo --enable-java-awt=gtk --enable-java-home --enable-languages=c,ada,c++,java,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 --target=aarch64-linux-gnu --with-arch-directory=aarch64 --with-default-libstdcxx-abi=new -v Processor Details- Kernel 3.14: Scaling Governor: meson_cpufreq interactive- Kernel 4.11: Scaling Governor: arm-big-little performance

Amlogic kernel 4.11 vs 3.14 on S905ramspeed: Integer Addramspeed: Integer Copyramspeed: Integer Scaleramspeed: Floating-Point Addstream: Addstream: Copystream: Scalecachebench: Read Cachecachebench: Write CacheKernel 3.14Kernel 4.113791.164502.624477.703790.774052.104133.924192.482160.053377.584149.734715.154661.494005.093619.953945.704160.122161.483379.42OpenBenchmarking.org

RAMspeed SMP

Integer Add

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Integer AddKernel 3.14Kernel 4.1190018002700360045003791.164149.73

RAMspeed SMP

Integer Copy

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Integer CopyKernel 3.14Kernel 4.11100020003000400050004502.624715.15

RAMspeed SMP

Integer Scale

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Integer ScaleKernel 3.14Kernel 4.11100020003000400050004477.704661.49

RAMspeed SMP

Floating-Point Add

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Floating-Point AddKernel 3.14Kernel 4.1190018002700360045003790.774005.09

Stream

Add

OpenBenchmarking.orgMB/s, More Is BetterStream 2013-01-17AddKernel 3.14Kernel 4.119001800270036004500SE +/- 6.14, N = 5SE +/- 46.14, N = 104052.103619.95-fipa-pta -march=armv8-a+crc1. (CC) gcc options: -Ofast -mcpu=cortex-a53 -fomit-frame-pointer -ftree-vectorize -fopenmp

Stream

Copy

OpenBenchmarking.orgMB/s, More Is BetterStream 2013-01-17CopyKernel 3.14Kernel 4.119001800270036004500SE +/- 11.56, N = 5SE +/- 37.23, N = 54133.923945.70-fipa-pta -march=armv8-a+crc1. (CC) gcc options: -Ofast -mcpu=cortex-a53 -fomit-frame-pointer -ftree-vectorize -fopenmp

Stream

Scale

OpenBenchmarking.orgMB/s, More Is BetterStream 2013-01-17ScaleKernel 3.14Kernel 4.119001800270036004500SE +/- 13.07, N = 5SE +/- 56.88, N = 54192.484160.12-fipa-pta -march=armv8-a+crc1. (CC) gcc options: -Ofast -mcpu=cortex-a53 -fomit-frame-pointer -ftree-vectorize -fopenmp

CacheBench

Read Cache

OpenBenchmarking.orgMB/s, More Is BetterCacheBenchRead CacheKernel 3.14Kernel 4.115001000150020002500SE +/- 1.15, N = 3SE +/- 0.04, N = 32160.052161.481. (CC) gcc options: -lrt

CacheBench

Write Cache

OpenBenchmarking.orgMB/s, More Is BetterCacheBenchWrite CacheKernel 3.14Kernel 4.117001400210028003500SE +/- 2.14, N = 3SE +/- 0.03, N = 33377.583379.421. (CC) gcc options: -lrt

Phoronix Test Suite v10.8.4