GCC Clang Intel Skylake Compiler Perf

GCC vs. LLVM Clang on Intel Core i5 Skylake system. Benchmarks by Michael Larabel for a future article on Phoronix.

HTML result view exported from: https://openbenchmarking.org/result/2103154-HA-1509040HA49.

GCC Clang Intel Skylake Compiler PerfProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerDisplay DriverOpenGLCompilerFile-SystemScreen ResolutionGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNClang 3.8 SVN OMPSR740_1509040-HA-GCCINTELS_compareIntel Core i5-6600K @ 3.50GHz (4 Cores)MSI Z170A GAMING PRO (MS-7984) v1.0Intel Sky Lake15360MB256GB TS256GSSD370SIntel Sky Lake (1150MHz)Realtek ALC1150DELL P2210HIntel ConnectionUbuntu 15.104.2.0-999-generic (x86_64) 20150901Xfce 4.12X Server 1.17.2intel 2.99.9173.3 Mesa 10.6.5GCC 4.9.3ext41920x1080GCC 5.2.0GCC 6.0.0 20150830Clang 3.8.0 (SVN 246796) + LLVM 3.8.0svn2 x Intel Xeon Silver 4214R @ 3.50GHz (24 Cores / 48 Threads)Yunke China 0DY2X0 (2.10.0 BIOS)Intel Sky Lake-E DMI3 Registers6 x 32 GB DDR4-3200MT/s HMA84GR7DJR4N-XN16001GB PERC H730P Adp + 480GB PERC H730P AdpMatrox G200eW32 x Intel I350 + 2 x Intel 82599ES 10-Gigabit SFI/SFP+CentOS Linux 73.10.0-1160.15.2.el7.x86_64 (x86_64)GCC 4.8.5 20150623xfs1024x768OpenBenchmarking.orgKernel Details- GCC 4.9.3: i915.preliminary_hw_support=1- GCC 5.2.0: i915.preliminary_hw_support=1- GCC 6.0.0 20150830: i915.preliminary_hw_support=1- Clang 3.8 SVN: i915.preliminary_hw_support=1- Clang 3.8 SVN OMP: i915.preliminary_hw_support=1- SR740_1509040-HA-GCCINTELS_compare: Transparent Huge Pages: alwaysCompiler Details- GCC 4.9.3: --disable-multilib --enable-checking=release --enable-languages=c,c++- GCC 5.2.0: --disable-multilib --enable-checking=release --enable-languages=c,c++- GCC 6.0.0 20150830: --disable-multilib --enable-checking=release --enable-languages=c,c++- Clang 3.8 SVN: Optimized build; Built Sep 3 2015 (15:42:57); Default target: x86_64-unknown-linux-gnu; Host CPU: skylake- Clang 3.8 SVN OMP: Optimized build; Built Sep 3 2015 (15:42:57); Default target: x86_64-unknown-linux-gnu; Host CPU: skylake- SR740_1509040-HA-GCCINTELS_compare: --build=x86_64-redhat-linux --disable-libgcj --disable-libunwind-exceptions --enable-__cxa_atexit --enable-bootstrap --enable-checking=release --enable-gnu-indirect-function --enable-gnu-unique-object --enable-initfini-array --enable-languages=c,c++,objc,obj-c++,java,fortran,ada,go,lto --enable-plugin --enable-shared --enable-threads=posix --mandir=/usr/share/man --with-arch_32=x86-64 --with-linker-hash-style=gnu --with-tune=generic Processor Details- GCC 4.9.3: Scaling Governor: acpi-cpufreq ondemand- GCC 5.2.0: Scaling Governor: acpi-cpufreq ondemand- GCC 6.0.0 20150830: Scaling Governor: acpi-cpufreq ondemand- Clang 3.8 SVN: Scaling Governor: acpi-cpufreq ondemand- Clang 3.8 SVN OMP: Scaling Governor: acpi-cpufreq ondemand- SR740_1509040-HA-GCCINTELS_compare: Scaling Governor: intel_pstate performance - CPU Microcode: 0x5003006Security Details- SR740_1509040-HA-GCCINTELS_compare: SELinux + itlb_multihit: KVM: Mitigation of Split huge pages + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of Load fences usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full retpoline IBPB + srbds: Not affected + tsx_async_abort: Mitigation of Clear buffers; SMT vulnerable

GCC Clang Intel Skylake Compiler Perfrodinia: OpenMP LavaMDrodinia: OpenMP CFD Solverrodinia: OpenMP Streamclusterhmmer: Pfam Database Searchscimark2: Compositescimark2: Monte Carloscimark2: Fast Fourier Transformscimark2: Sparse Matrix Multiplyscimark2: Dense LU Matrix Factorizationscimark2: Jacobi Successive Over-Relaxationjohn-the-ripper: Blowfishjohn-the-ripper: Traditional DESjohn-the-ripper: MD5x264: H.264 Video Encodinghimeno: Poisson Pressure Solverc-ray: Total Timesmallpt: Global Illumination Renderer; 100 Samplesencode-flac: WAV To FLACencode-mp3: WAV To MP3hint: FLOATapache: Static Web Page ServingGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNClang 3.8 SVN OMPSR740_1509040-HA-GCCINTELS_compare215.6765.9835.279.331594.66743.37373.002739.392992.341125.2047552448400076544169.172763.0715.17284.9310.61394717206.9144411.00216.2065.1835.019.161594.01754.86372.762749.082968.231125.1147602502533377004171.092744.7615.11284.9710.77402676204.8041662.12215.7263.2735.009.241584.33750.18374.272693.952979.521125.2347822514133376902171.242720.0015.09284.9410.56404526879.2343039.62856.60183.97128.849.522467.10644.15383.762721.337111.791474.451224629133316684167.171976.5734.591235.1411.33293947985.3144513.53857.02184.63128.2012246283667166703148.96217.54421.7196.6221478.45653.74430.132429.502875.291003.6034974111214333442368486.292332.6645964.2021157.250361326865.9790513327.92OpenBenchmarking.org

Rodinia

Test: OpenMP LavaMD

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 2.4Test: OpenMP LavaMDGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNClang 3.8 SVN OMPSR740_1509040-HA-GCCINTELS_compare2004006008001000SE +/- 0.20, N = 3SE +/- 0.09, N = 3SE +/- 0.14, N = 3SE +/- 0.10, N = 3SE +/- 0.35, N = 3SE +/- 0.09, N = 3215.67216.20215.72856.60857.0248.961. (CXX) g++ options: -O2 -lOpenCL

Rodinia

Test: OpenMP CFD Solver

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 2.4Test: OpenMP CFD SolverGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNClang 3.8 SVN OMPSR740_1509040-HA-GCCINTELS_compare4080120160200SE +/- 0.21, N = 3SE +/- 0.19, N = 3SE +/- 0.63, N = 3SE +/- 0.19, N = 3SE +/- 0.62, N = 3SE +/- 0.11, N = 365.9865.1863.27183.97184.6317.541. (CXX) g++ options: -O2 -lOpenCL

Rodinia

Test: OpenMP Streamcluster

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 2.4Test: OpenMP StreamclusterGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNClang 3.8 SVN OMPSR740_1509040-HA-GCCINTELS_compare306090120150SE +/- 0.04, N = 3SE +/- 0.05, N = 3SE +/- 0.01, N = 3SE +/- 0.58, N = 3SE +/- 0.06, N = 3SE +/- 0.60, N = 1535.2735.0135.00128.84128.2021.721. (CXX) g++ options: -O2 -lOpenCL

Timed HMMer Search

Pfam Database Search

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed HMMer Search 2.3.2Pfam Database SearchGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNSR740_1509040-HA-GCCINTELS_compare3691215SE +/- 0.047, N = 3SE +/- 0.003, N = 3SE +/- 0.056, N = 3SE +/- 0.031, N = 3SE +/- 0.019, N = 39.3309.1609.2409.5206.622-O3 -march=native-O3 -march=native-O3 -march=native-O3 -march=native-O21. (CC) gcc options: -pthread -lhmmer -lsquid -lm

SciMark

Computational Test: Composite

OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: CompositeGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNSR740_1509040-HA-GCCINTELS_compare5001000150020002500SE +/- 0.46, N = 4SE +/- 1.66, N = 4SE +/- 0.86, N = 4SE +/- 1.28, N = 4SE +/- 5.81, N = 41594.661594.011584.332467.101478.45-O3 -march=native-O3 -march=native-O3 -march=native-O3 -march=native1. (CXX) g++ options:

SciMark

Computational Test: Monte Carlo

OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Monte CarloGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNSR740_1509040-HA-GCCINTELS_compare160320480640800SE +/- 0.01, N = 4SE +/- 0.02, N = 4SE +/- 0.00, N = 4SE +/- 0.05, N = 4SE +/- 1.90, N = 4743.37754.86750.18644.15653.74-O3 -march=native-O3 -march=native-O3 -march=native-O3 -march=native1. (CXX) g++ options:

SciMark

Computational Test: Fast Fourier Transform

OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Fast Fourier TransformGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNSR740_1509040-HA-GCCINTELS_compare90180270360450SE +/- 1.15, N = 4SE +/- 0.57, N = 4SE +/- 0.67, N = 4SE +/- 0.82, N = 4SE +/- 4.85, N = 4373.00372.76374.27383.76430.13-O3 -march=native-O3 -march=native-O3 -march=native-O3 -march=native1. (CXX) g++ options:

SciMark

Computational Test: Sparse Matrix Multiply

OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Sparse Matrix MultiplyGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNSR740_1509040-HA-GCCINTELS_compare6001200180024003000SE +/- 2.49, N = 4SE +/- 1.46, N = 4SE +/- 2.13, N = 3SE +/- 2.25, N = 4SE +/- 19.09, N = 42739.392749.082693.952721.332429.50-O3 -march=native-O3 -march=native-O3 -march=native-O3 -march=native1. (CXX) g++ options:

SciMark

Computational Test: Dense LU Matrix Factorization

OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Dense LU Matrix FactorizationGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNSR740_1509040-HA-GCCINTELS_compare15003000450060007500SE +/- 0.38, N = 4SE +/- 7.45, N = 4SE +/- 2.56, N = 4SE +/- 5.00, N = 4SE +/- 12.94, N = 42992.342968.232979.527111.792875.29-O3 -march=native-O3 -march=native-O3 -march=native-O3 -march=native1. (CXX) g++ options:

SciMark

Computational Test: Jacobi Successive Over-Relaxation

OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Jacobi Successive Over-RelaxationGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNSR740_1509040-HA-GCCINTELS_compare30060090012001500SE +/- 0.12, N = 4SE +/- 0.03, N = 4SE +/- 0.01, N = 4SE +/- 0.06, N = 4SE +/- 3.88, N = 41125.201125.111125.231474.451003.60-O3 -march=native-O3 -march=native-O3 -march=native-O3 -march=native1. (CXX) g++ options:

John The Ripper

Test: Blowfish

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 1.8.0Test: BlowfishGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNClang 3.8 SVN OMPSR740_1509040-HA-GCCINTELS_compare7K14K21K28K35KSE +/- 3.00, N = 3SE +/- 4.10, N = 3SE +/- 1.67, N = 3SE +/- 0.67, N = 3SE +/- 1.86, N = 3SE +/- 72.22, N = 347554760478212241224349741. (CC) gcc options: -fopenmp -lcrypt

John The Ripper

Test: Traditional DES

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 1.8.0Test: Traditional DESGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNClang 3.8 SVN OMPSR740_1509040-HA-GCCINTELS_compare20M40M60M80M100MSE +/- 175214.15, N = 3SE +/- 72723.98, N = 3SE +/- 10397.65, N = 3SE +/- 2333.33, N = 3SE +/- 666.67, N = 3SE +/- 245045.80, N = 3244840002502533325141333629133362836671112143331. (CC) gcc options: -fopenmp -lcrypt

John The Ripper

Test: MD5

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 1.8.0Test: MD5GCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNClang 3.8 SVN OMPSR740_1509040-HA-GCCINTELS_compare90K180K270K360K450KSE +/- 51.00, N = 3SE +/- 0.00, N = 3SE +/- 51.00, N = 3SE +/- 0.33, N = 3SE +/- 10.67, N = 376544770047690216684166704423681. (CC) gcc options: -fopenmp -lcrypt

x264

H.264 Video Encoding

OpenBenchmarking.orgFrames Per Second, More Is Betterx264 2014-08-30H.264 Video EncodingGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNSR740_1509040-HA-GCCINTELS_compare110220330440550SE +/- 0.56, N = 5SE +/- 0.39, N = 5SE +/- 0.43, N = 5SE +/- 0.27, N = 5SE +/- 2.04, N = 5169.17171.09171.24167.17486.29-march=native-march=native-march=native-march=native1. (CC) gcc options: -ldl -m64 -lm -lpthread -O3 -ffast-math -std=gnu99 -fomit-frame-pointer -fno-tree-vectorize

Himeno Benchmark

Poisson Pressure Solver

OpenBenchmarking.orgMFLOPS, More Is BetterHimeno Benchmark 3.0Poisson Pressure SolverGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNSR740_1509040-HA-GCCINTELS_compare6001200180024003000SE +/- 0.74, N = 3SE +/- 14.56, N = 3SE +/- 3.11, N = 3SE +/- 1.58, N = 3SE +/- 8.51, N = 32763.072744.762720.001976.572332.66-march=native-march=native-march=native-march=native1. (CC) gcc options: -O3

C-Ray

Total Time

OpenBenchmarking.orgSeconds, Fewer Is BetterC-Ray 1.1Total TimeGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNSR740_1509040-HA-GCCINTELS_compare816243240SE +/- 0.012, N = 3SE +/- 0.002, N = 3SE +/- 0.003, N = 3SE +/- 0.010, N = 3SE +/- 0.014, N = 315.17015.11015.09034.5904.202-march=native-march=native-march=native-march=native1. (CC) gcc options: -lm -lpthread -O3

Smallpt

Global Illumination Renderer; 100 Samples

OpenBenchmarking.orgSeconds, Fewer Is BetterSmallpt 1.0Global Illumination Renderer; 100 SamplesGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNClang 3.8 SVN OMPSR740_1509040-HA-GCCINTELS_compare306090120150SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 12.87, N = 1228282812331115-O3 -march=native-O3 -march=native-O3 -march=native-O3 -march=native-O3 -march=native -fopenmp=libomp -fno-exceptions1. (CXX) g++ options: -fopenmp

FLAC Audio Encoding

WAV To FLAC

OpenBenchmarking.orgSeconds, Fewer Is BetterFLAC Audio Encoding 1.3.1WAV To FLACGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNSR740_1509040-HA-GCCINTELS_compare246810SE +/- 0.002, N = 5SE +/- 0.002, N = 5SE +/- 0.002, N = 5SE +/- 0.003, N = 5SE +/- 0.037, N = 54.9304.9704.9405.1407.250-O3 -march=native -fvisibility=hidden-O3 -march=native -fvisibility=hidden-O3 -march=native -fvisibility=hidden-O3 -march=native -logg-O2 -fvisibility=hidden1. (CXX) g++ options: -lm

LAME MP3 Encoding

WAV To MP3

OpenBenchmarking.orgSeconds, Fewer Is BetterLAME MP3 Encoding 3.99.3WAV To MP3GCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVN3691215SE +/- 0.00, N = 5SE +/- 0.00, N = 5SE +/- 0.00, N = 5SE +/- 0.00, N = 510.6110.7710.5611.331. (CC) gcc options: -pipe -O3 -march=native -lm

Hierarchical INTegration

Test: FLOAT

OpenBenchmarking.orgQUIPs, More Is BetterHierarchical INTegration 1.0Test: FLOATGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNSR740_1509040-HA-GCCINTELS_compare90M180M270M360M450MSE +/- 252594.71, N = 3SE +/- 279552.00, N = 3SE +/- 358895.73, N = 3SE +/- 129692.07, N = 3SE +/- 144132.51, N = 3394717206.91402676204.80404526879.23293947985.31361326865.981. (CC) gcc options: -O3 -march=native -lm

Apache Benchmark

Static Web Page Serving

OpenBenchmarking.orgRequests Per Second, More Is BetterApache Benchmark 2.4.7Static Web Page ServingGCC 4.9.3GCC 5.2.0GCC 6.0.0 20150830Clang 3.8 SVNSR740_1509040-HA-GCCINTELS_compare10K20K30K40K50KSE +/- 154.80, N = 3SE +/- 288.36, N = 3SE +/- 1006.67, N = 6SE +/- 388.92, N = 3SE +/- 147.37, N = 344411.0041662.1243039.6244513.5313327.92-O3 -march=native-O3 -march=native-O3 -march=native-O3 -march=native-O21. (CC) gcc options: -shared -fPIC -pthread

Phoronix Test Suite v10.8.4