GCC AMD Bulldozer Compiler Tuning

AMD FX-8150 Bulldozer compiler tuning using GCC 4.7.1 with different march options for the test profiles to look at the latest AMD GCC performance. Benchmarking by Michael Larabel for a future article on Phoronix.com.

HTML result view exported from: https://openbenchmarking.org/result/1209227-RA-GCCAMDBUL32&grs&rdt.

GCC AMD Bulldozer Compiler TuningProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerDisplay DriverOpenGLCompilerFile-SystemScreen Resolutionmarch=k8march=bdver1march=barcelonamarch=noconamarch=core2march=k8-sse3AMD FX-8150 Eight-Core @ 3.60GHz (8 Cores)ASUS Crosshair V FormulaAMD ATI RD890 bridge4096MB60GB OCZ VERTEX2NVIDIA GeForce 9600 GSO 512MB (399/399MHz)Realtek ALC889DELL P2210HIntel 82583V Gigabit ConnectionUbuntu 12.103.5.0-14-generic (x86_64)Unity 6.4.0X Server 1.13.0nouveau 1.0.23.0 Mesa 8.1-devel Gallium 0.4GCC 4.7ext41920x1080OpenBenchmarking.orgCompiler Details- --build=x86_64-linux-gnu --disable-werror --enable-checking=release --enable-clocale=gnu --enable-gnu-unique-object --enable-languages=c,c++,go,fortran,objc,obj-c++ --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-nls --enable-objc-gc --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --target=x86_64-linux-gnu --with-arch-32=i686 --with-tune=generic -v Processor Details- Scaling Governor: ondemandSystem Details- march=k8: Compiz and Firefox were running on this system.- march=bdver1: Compiz and Firefox were running on this system.- march=barcelona: Compiz was running on this system.- march=nocona: Compiz was running on this system.- march=core2: Compiz was running on this system.- march=k8-sse3: Compiz was running on this system.

GCC AMD Bulldozer Compiler Tuningc-ray: Total Timegraphics-magick: Sharpensmallpt: Global Illumination Renderer; 100 Samplesnero2d: Total Timegraphics-magick: Blurgraphics-magick: Resizinghimeno: Poisson Pressure Solvergraphics-magick: Local Adaptive Thresholdingbuild-php: Time To Compilepgbench: TPC-B Transactions Per Secondgraphics-magick: HWB Color Spacemarch=k8march=bdver1march=barcelonamarch=noconamarch=core2march=k8-sse352.846344643.5997128614.836733.021986.1815126.9910030561.85113144698.926634.411916.5115736.098632534.42107139669.286133.072037.1115636.846533549.4399126654.276831.202064.4016036.389933531.2796139670.756733.222039.7115852.826244636.9995126678.976632.982018.34150OpenBenchmarking.org

C-Ray

Total Time

OpenBenchmarking.orgSeconds, Fewer Is BetterC-Ray 1.1Total Timemarch=k8march=bdver1march=barcelonamarch=noconamarch=core2march=k8-sse31224364860SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 352.8426.9936.0936.8436.3852.82-march=k8-march=bdver1-march=barcelona-march=nocona-march=core2-march=k8-sse31. (CC) gcc options: -lm -lpthread -O3

GraphicsMagick

Operation: Sharpen

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.12Operation: Sharpenmarch=k8march=bdver1march=barcelonamarch=noconamarch=core2march=k8-sse320406080100SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 36310086659962-march=k8-march=bdver1-march=barcelona-march=nocona-march=core2-march=k8-sse31. (CC) gcc options: -std=gnu99 -fopenmp -O3 -pthread -lz -lm -lgomp -lpthread

Smallpt

Global Illumination Renderer; 100 Samples

OpenBenchmarking.orgSeconds, Fewer Is BetterSmallpt 1.0Global Illumination Renderer; 100 Samplesmarch=k8march=bdver1march=barcelonamarch=noconamarch=core2march=k8-sse31020304050SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3443032333344-march=k8-march=bdver1-march=barcelona-march=nocona-march=core2-march=k8-sse31. (CXX) g++ options: -fopenmp -O3

Open FMM Nero2D

Total Time

OpenBenchmarking.orgSeconds, Fewer Is BetterOpen FMM Nero2D 2.0.2Total Timemarch=k8march=bdver1march=barcelonamarch=noconamarch=core2march=k8-sse3140280420560700643.59561.85534.42549.43531.27636.99-march=k8-march=bdver1-march=barcelona-march=nocona-march=core2-march=k8-sse31. (CXX) g++ options: -O3 -lfftw3 -llapack -lblas -lgfortran -lquadmath -lm

GraphicsMagick

Operation: Blur

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.12Operation: Blurmarch=k8march=bdver1march=barcelonamarch=noconamarch=core2march=k8-sse3306090120150SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 1.15, N = 3SE +/- 0.67, N = 3SE +/- 0.00, N = 3SE +/- 0.58, N = 397113107999695-march=k8-march=bdver1-march=barcelona-march=nocona-march=core2-march=k8-sse31. (CC) gcc options: -std=gnu99 -fopenmp -O3 -pthread -lz -lm -lgomp -lpthread

GraphicsMagick

Operation: Resizing

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.12Operation: Resizingmarch=k8march=bdver1march=barcelonamarch=noconamarch=core2march=k8-sse3306090120150SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3128144139126139126-march=k8-march=bdver1-march=barcelona-march=nocona-march=core2-march=k8-sse31. (CC) gcc options: -std=gnu99 -fopenmp -O3 -pthread -lz -lm -lgomp -lpthread

Himeno Benchmark

Poisson Pressure Solver

OpenBenchmarking.orgMFLOPS, More Is BetterHimeno Benchmark 3.0Poisson Pressure Solvermarch=k8march=bdver1march=barcelonamarch=noconamarch=core2march=k8-sse3150300450600750SE +/- 0.75, N = 3SE +/- 8.36, N = 3SE +/- 6.64, N = 3SE +/- 5.51, N = 3SE +/- 8.14, N = 3SE +/- 7.84, N = 3614.83698.92669.28654.27670.75678.97-march=k8-march=bdver1-march=barcelona-march=nocona-march=core2-march=k8-sse31. (CC) gcc options: -O3

GraphicsMagick

Operation: Local Adaptive Thresholding

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.12Operation: Local Adaptive Thresholdingmarch=k8march=bdver1march=barcelonamarch=noconamarch=core2march=k8-sse31530456075SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3676661686766-march=k8-march=bdver1-march=barcelona-march=nocona-march=core2-march=k8-sse31. (CC) gcc options: -std=gnu99 -fopenmp -O3 -pthread -lz -lm -lgomp -lpthread

Timed PHP Compilation

Time To Compile

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed PHP Compilation 5.2.9Time To Compilemarch=k8march=bdver1march=barcelonamarch=noconamarch=core2march=k8-sse3816243240SE +/- 0.06, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.01, N = 3SE +/- 0.05, N = 3SE +/- 0.09, N = 333.0234.4133.0731.2033.2232.98-march=k8-march=bdver1-march=barcelona-march=nocona-march=core2-march=k8-sse31. (CC) gcc options: -O3 -pedantic -ldl -lz -lm

PostgreSQL pgbench

TPC-B Transactions Per Second

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 8.4.11TPC-B Transactions Per Secondmarch=k8march=bdver1march=barcelonamarch=noconamarch=core2march=k8-sse3400800120016002000SE +/- 40.09, N = 6SE +/- 33.50, N = 6SE +/- 1.46, N = 3SE +/- 18.47, N = 3SE +/- 33.91, N = 3SE +/- 8.45, N = 31986.181916.512037.112064.402039.712018.34-march=k8-march=bdver1-march=barcelona-march=nocona-march=core2-march=k8-sse31. (CC) gcc options: -O3 -fno-strict-aliasing -fwrapv -lpgport -lpq -lcrypt -ldl -lm

GraphicsMagick

Operation: HWB Color Space

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.12Operation: HWB Color Spacemarch=k8march=bdver1march=barcelonamarch=noconamarch=core2march=k8-sse34080120160200SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.67, N = 3SE +/- 1.20, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3151157156160158150-march=k8-march=bdver1-march=barcelona-march=nocona-march=core2-march=k8-sse31. (CC) gcc options: -std=gnu99 -fopenmp -O3 -pthread -lz -lm -lgomp -lpthread


Phoronix Test Suite v10.8.4