Linux 4.7 CPUFreq Schedutil Testing

Linux 4.7 kernel benchmarking. Tests by Michael Larabel for a future article.

HTML result view exported from: https://openbenchmarking.org/result/1605180-HA-LINUX47CP71&grt&sro.

Linux 4.7 CPUFreq Schedutil TestingProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerDisplay DriverOpenGLCompilerFile-SystemScreen ResolutionCPUFreq SchedutilCPUFreq OndemandCPUFreq ConservativeCPUFreq: PowersaveCPUFreq PerformanceP-State PowersaveP-State PerformanceIntel Xeon E5-2687W v3 @ 3.10GHz (20 Cores)MSI X99S SLI PLUS (MS-7885) v1.0Intel Xeon E7 v3/Xeon16384MBPNY CS1211 120GB + 80GB INTEL SSDSCKGW08AMD FirePro V7900 2048MBRealtek ALC892ASUS PB278Intel ConnectionUbuntu 16.044.6.0-phx-schedutil (x86_64)Unity 7.4.0X Server 1.18.3modesetting 1.18.34.1 Mesa 11.2.0 Gallium 0.4GCC 5.3.1 20160413ext42560x1440Intel Xeon E5-2687W v3 @ 3.50GHz (20 Cores)OpenBenchmarking.orgCompiler Details- --build=x86_64-linux-gnu --disable-browser-plugin --disable-vtable-verify --disable-werror --enable-checking=release --enable-clocale=gnu --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-libmpx --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-arch-directory=amd64 --with-default-libstdcxx-abi=new --with-multilib-list=m32,m64,mx32 --with-tune=generic -v Processor Details- CPUFreq Schedutil: Scaling Governor: acpi-cpufreq schedutil- CPUFreq Ondemand: Scaling Governor: acpi-cpufreq ondemand- CPUFreq Conservative: Scaling Governor: acpi-cpufreq conservative- CPUFreq: Powersave: Scaling Governor: acpi-cpufreq powersave- CPUFreq Performance: Scaling Governor: acpi-cpufreq performance- P-State Powersave: Scaling Governor: intel_pstate powersave- P-State Performance: Scaling Governor: intel_pstate performanceGraphics Details- EXA

Linux 4.7 CPUFreq Schedutil Testingapache: Static Web Page Servingc-ray: Total Timeclomp: Static OMP Speedupffte: N=64, 1D Complex FFT Routineencode-flac: WAV To FLACgraphics-magick: Resizinghimeno: Poisson Pressure Solverhpcc: G-HPLjohn-the-ripper: Blowfishencode-mp3: WAV To MP3lammps: Rhodopsin Proteinmultichase: 256MB Array, 256 Byte Stridemultichase: 1GB Array, 256 Byte Stride, 4 Threadsnpb: EP.Bopm-git: Upscale-Relperm - 16pgbench: Buffer Test - Heavy Contention - Read Writeredis: GETtesseract: 2560 x 1440build-linux-kernel: Time To Compilex264: H.264 Video Encodingxonotic: 2560 x 1440 - LowCPUFreq SchedutilCPUFreq OndemandCPUFreq ConservativeCPUFreq: PowersaveCPUFreq PerformanceP-State PowersaveP-State Performance20379.4610.776.404965.767.691781732.81103.823671413613.8940.3065.2474.95368.2913.395273.80606865.1463.6654.31350.6059.3822338.8810.776.405276.476.921951772.83103.778001411212.6040.0668.1972.74370.1813.505315.35615364.5864.3454.47309.9559.1221548.4110.906.595232.437.051631749.26103.937001412612.8640.1564.7774.35360.8013.764695.58526209.7463.7856.25316.8759.3613062.7928.716.951814.0618.9270624.4540.40830528137.55113.48106.10110.97139.0926.614905.52379283.7663.44135.35142.9459.3324517.7010.736.424991.636.901981748.54103.868001411712.9741.0068.0871.93370.3513.565327.96609017.6263.6752.43369.2759.1213282.1629.576.111814.7018.9568625.0231.57813458437.29113.48107.03112.77111.974852.10387649.3163.71136.33136.0259.6513077.7629.596.424988.7418.9368623.92104.851671413637.41112.4469.4673.16105.234832.19387296.9463.29135.89134.5359.26OpenBenchmarking.org

Apache Benchmark

Static Web Page Serving

OpenBenchmarking.orgRequests Per Second, More Is BetterApache Benchmark 2.4.7Static Web Page ServingCPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State PerformanceP-State Powersave5K10K15K20K25KSE +/- 982.79, N = 3SE +/- 117.89, N = 3SE +/- 183.42, N = 3SE +/- 82.65, N = 3SE +/- 9.66, N = 3SE +/- 73.74, N = 3SE +/- 31.83, N = 321548.4122338.8824517.7020379.4613062.7913077.7613282.161. (CC) gcc options: -shared -fPIC -O2 -pthread

C-Ray

Total Time

OpenBenchmarking.orgSeconds, Fewer Is BetterC-Ray 1.1Total TimeCPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State PerformanceP-State Powersave714212835SE +/- 0.14, N = 3SE +/- 0.03, N = 3SE +/- 0.04, N = 3SE +/- 0.03, N = 3SE +/- 0.06, N = 3SE +/- 0.05, N = 3SE +/- 0.06, N = 310.9010.7710.7310.7728.7129.5929.571. (CC) gcc options: -lm -lpthread -O3

CLOMP

Static OMP Speedup

OpenBenchmarking.orgSpeedup, More Is BetterCLOMP 3.3Static OMP SpeedupCPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State PerformanceP-State Powersave246810SE +/- 0.23, N = 5SE +/- 0.10, N = 5SE +/- 0.10, N = 5SE +/- 0.08, N = 5SE +/- 0.05, N = 5SE +/- 0.11, N = 5SE +/- 0.85, N = 56.596.406.426.406.956.426.111. (CC) gcc options: --openmp -O3 -lm

CPU Frequency (CPU0) Monitor

Phoronix Test Suite System Monitoring

OpenBenchmarking.orgMegahertzCPU Frequency (CPU0) MonitorPhoronix Test Suite System MonitoringCPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State Performance6001200180024003000Min: 1200 / Avg: 2249.66 / Max: 3101Min: 1200 / Avg: 2318.98 / Max: 3101Min: 3101 / Avg: 3101 / Max: 3101Min: 1200 / Avg: 2475.16 / Max: 3101Min: 1200 / Avg: 1200 / Max: 1200Min: 0 / Avg: 1512.39 / Max: 3500.93

FFTE

Test: N=64, 1D Complex FFT Routine

OpenBenchmarking.orgMFLOPS, More Is BetterFFTE 5.0Test: N=64, 1D Complex FFT RoutineCPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State PerformanceP-State Powersave11002200330044005500SE +/- 46.09, N = 3SE +/- 3.96, N = 3SE +/- 161.87, N = 3SE +/- 82.32, N = 3SE +/- 0.04, N = 3SE +/- 146.87, N = 3SE +/- 0.06, N = 35232.435276.474991.634965.761814.064988.741814.701. (F9X) gfortran options: -O3 -fomit-frame-pointer -fopenmp -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

FLAC Audio Encoding

WAV To FLAC

OpenBenchmarking.orgSeconds, Fewer Is BetterFLAC Audio Encoding 1.3.1WAV To FLACCPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State PerformanceP-State Powersave510152025SE +/- 0.07, N = 5SE +/- 0.01, N = 5SE +/- 0.00, N = 5SE +/- 0.46, N = 5SE +/- 0.01, N = 5SE +/- 0.01, N = 5SE +/- 0.01, N = 57.056.926.907.6918.9218.9318.951. (CXX) g++ options: -O2 -fvisibility=hidden -lm

GraphicsMagick

Operation: Resizing

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.19Operation: ResizingCPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State PerformanceP-State Powersave4080120160200SE +/- 0.88, N = 3SE +/- 0.67, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 31631951981787068681. (CC) gcc options: -fopenmp -O2 -pthread -lXext -lSM -lICE -lX11 -lz -lm -lgomp -lpthread

Himeno Benchmark

Poisson Pressure Solver

OpenBenchmarking.orgMFLOPS, More Is BetterHimeno Benchmark 3.0Poisson Pressure SolverCPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State PerformanceP-State Powersave400800120016002000SE +/- 9.47, N = 3SE +/- 2.65, N = 3SE +/- 7.56, N = 3SE +/- 7.34, N = 3SE +/- 0.20, N = 3SE +/- 0.44, N = 3SE +/- 0.81, N = 31749.261772.831748.541732.81624.45623.92625.021. (CC) gcc options: -O3 -mavx2

HPC Challenge

Test / Class: G-HPL

OpenBenchmarking.orgGFLOPS, More Is BetterHPC Challenge 1.4.3Test / Class: G-HPLCPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State PerformanceP-State Powersave20406080100SE +/- 0.05, N = 3SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.05, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.40, N = 3103.94103.78103.87103.8240.41104.8531.581. (CC) gcc options: -lblas -lm -pthread -lmpi -fomit-frame-pointer -O3 -march=native -funroll-loops2. BLAS + Open MPI 1.10.2

John The Ripper

Test: Blowfish

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 1.8.0Test: BlowfishCPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State PerformanceP-State Powersave3K6K9K12K15KSE +/- 9.67, N = 3SE +/- 0.00, N = 3SE +/- 9.67, N = 3SE +/- 0.00, N = 3SE +/- 6.94, N = 3SE +/- 0.00, N = 3SE +/- 411.91, N = 31412614112141171413652811413645841. (CC) gcc options: -fopenmp -lcrypt

LAME MP3 Encoding

WAV To MP3

OpenBenchmarking.orgSeconds, Fewer Is BetterLAME MP3 Encoding 3.99.3WAV To MP3CPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State PerformanceP-State Powersave918273645SE +/- 0.14, N = 5SE +/- 0.02, N = 5SE +/- 0.04, N = 5SE +/- 0.54, N = 5SE +/- 0.17, N = 5SE +/- 0.12, N = 5SE +/- 0.21, N = 512.8612.6012.9713.8937.5537.4137.291. (CC) gcc options: -O3 -ffast-math -funroll-loops -pipe -lncurses -lm

LAMMPS Molecular Dynamics Simulator

Test: Rhodopsin Protein

OpenBenchmarking.orgLoop Time, Fewer Is BetterLAMMPS Molecular Dynamics Simulator 1.0Test: Rhodopsin ProteinCPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State PerformanceP-State Powersave306090120150SE +/- 0.12, N = 3SE +/- 0.27, N = 3SE +/- 0.17, N = 3SE +/- 0.33, N = 3SE +/- 0.26, N = 3SE +/- 0.32, N = 3SE +/- 0.21, N = 340.1540.0641.0040.30113.48112.44113.481. (CXX) g++ options: -lfftw -lmpich

Multichase Pointer Chaser

Test: 256MB Array, 256 Byte Stride

OpenBenchmarking.orgns, Fewer Is BetterMultichase Pointer ChaserTest: 256MB Array, 256 Byte StrideCPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State PerformanceP-State Powersave20406080100SE +/- 2.80, N = 3SE +/- 1.23, N = 3SE +/- 1.02, N = 3SE +/- 2.29, N = 3SE +/- 0.81, N = 3SE +/- 0.05, N = 3SE +/- 0.27, N = 364.7768.1968.0865.24106.1069.46107.031. (CC) gcc options: -O2 -static -pthread -lrt

Multichase Pointer Chaser

Test: 1GB Array, 256 Byte Stride, 4 Threads

OpenBenchmarking.orgns, Fewer Is BetterMultichase Pointer ChaserTest: 1GB Array, 256 Byte Stride, 4 ThreadsCPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State PerformanceP-State Powersave306090120150SE +/- 0.27, N = 3SE +/- 0.92, N = 3SE +/- 1.60, N = 3SE +/- 0.13, N = 3SE +/- 1.09, N = 3SE +/- 0.64, N = 3SE +/- 0.12, N = 374.3572.7471.9374.95110.9773.16112.771. (CC) gcc options: -O2 -static -pthread -lrt

NAS Parallel Benchmarks

Test / Class: EP.B

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.3Test / Class: EP.BCPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State PerformanceP-State Powersave80160240320400SE +/- 5.17, N = 3SE +/- 1.48, N = 3SE +/- 1.54, N = 3SE +/- 2.78, N = 3SE +/- 0.74, N = 3SE +/- 8.12, N = 3SE +/- 8.66, N = 3360.80370.18370.35368.29139.09105.23111.971. (F9X) gfortran options: -fopenmp

Open Porous Media Git

OPM Benchmark: Upscale-Relperm - Threads: 16

OpenBenchmarking.orgSeconds, Fewer Is BetterOpen Porous Media GitOPM Benchmark: Upscale-Relperm - Threads: 16CPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: Powersave612182430SE +/- 0.13, N = 3SE +/- 0.22, N = 3SE +/- 0.20, N = 3SE +/- 0.20, N = 3SE +/- 0.21, N = 313.7613.5013.5613.3926.611. Build Time Tue May 17 12:05:40 EDT 2016;

PostgreSQL pgbench

Scaling: Buffer Test - Test: Heavy Contention - Mode: Read Write

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 9.4.3Scaling: Buffer Test - Test: Heavy Contention - Mode: Read WriteCPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State PerformanceP-State Powersave11002200330044005500SE +/- 205.11, N = 3SE +/- 8.74, N = 3SE +/- 302.93, N = 3SE +/- 50.60, N = 3SE +/- 5.76, N = 3SE +/- 4.75, N = 3SE +/- 9.49, N = 34695.585315.355327.965273.804905.524832.194852.101. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -pthread -lpgcommon -lpgport -lpq -lpthread -lrt -lcrypt -ldl -lm

Redis

Test: GET

OpenBenchmarking.orgRequests Per Second, More Is BetterRedis 3.0.1Test: GETCPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State PerformanceP-State Powersave130K260K390K520K650KSE +/- 9602.66, N = 3SE +/- 10591.07, N = 3SE +/- 1131.35, N = 3SE +/- 4579.25, N = 3SE +/- 3493.78, N = 3SE +/- 229.02, N = 3SE +/- 674.73, N = 3526209.74615364.58609017.62606865.14379283.76387296.94387649.311. (CC) gcc options: -ggdb -rdynamic -lm -pthread -ldl -std=gnu99 -pipe -g3 -O3 -funroll-loops

Tesseract

Resolution: 2560 x 1440

OpenBenchmarking.orgFrames Per Second, More Is BetterTesseract 2014-05-12Resolution: 2560 x 1440CPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State PerformanceP-State Powersave1428425670SE +/- 0.31, N = 3SE +/- 0.50, N = 3SE +/- 0.09, N = 3SE +/- 0.23, N = 3SE +/- 0.05, N = 3SE +/- 0.16, N = 3SE +/- 0.26, N = 363.7864.3463.6763.6663.4463.2963.71

Timed Linux Kernel Compilation

Time To Compile

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 4.3Time To CompileCPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State PerformanceP-State Powersave306090120150SE +/- 0.74, N = 3SE +/- 0.88, N = 3SE +/- 0.80, N = 3SE +/- 1.35, N = 3SE +/- 2.43, N = 3SE +/- 2.42, N = 3SE +/- 1.92, N = 356.2554.4752.4354.31135.35135.89136.33

x264

H.264 Video Encoding

OpenBenchmarking.orgFrames Per Second, More Is Betterx264 2015-11-02H.264 Video EncodingCPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State PerformanceP-State Powersave80160240320400SE +/- 27.47, N = 5SE +/- 1.38, N = 5SE +/- 2.51, N = 5SE +/- 2.84, N = 5SE +/- 0.68, N = 5SE +/- 0.56, N = 5SE +/- 0.57, N = 5316.87309.95369.27350.60142.94134.53136.021. (CC) gcc options: -ldl -m64 -lm -lpthread -O3 -ffast-math -std=gnu99 -fomit-frame-pointer -fno-tree-vectorize

Xonotic

Resolution: 2560 x 1440 - Effects Quality: Low

OpenBenchmarking.orgFrames Per Second, More Is BetterXonotic 0.8Resolution: 2560 x 1440 - Effects Quality: LowCPUFreq ConservativeCPUFreq OndemandCPUFreq PerformanceCPUFreq SchedutilCPUFreq: PowersaveP-State PerformanceP-State Powersave1326395265SE +/- 0.18, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.14, N = 3SE +/- 0.00, N = 3SE +/- 0.07, N = 3SE +/- 0.12, N = 359.3659.1259.1259.3859.3359.2659.65MIN: 42 / MAX: 61MIN: 41 / MAX: 60MIN: 42 / MAX: 60MIN: 42 / MAX: 61MIN: 41 / MAX: 61MIN: 42 / MAX: 62MIN: 42 / MAX: 61


Phoronix Test Suite v10.8.4