LLVM Clang 3.3 vs. GCC 4.8 - Intel Core-AVX2 Haswell

Intel Core i7-4770K testing with a Intel DH87RL motherboard looking at the GCC 4.7, GCC 4.8, LLVM Clang 3.2, and LLVM Clang 3.3 compiler performance with core-avx2 Haswell optimizations. Intel Core i7 Haswell core-avx2 compiler benchmarks for a future article on Phoronix by Michael Larabel.

HTML result view exported from: https://openbenchmarking.org/result/1306206-SO-LLVMCLANG08&grr&sor.

LLVM Clang 3.3 vs. GCC 4.8 - Intel Core-AVX2 HaswellProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerDisplay DriverOpenGLCompilerFile-SystemScreen ResolutionGCC 4.7.3GCC 4.8.1LLVM Clang 3.2LLVM Clang 3.3Intel Core i7-4770K @ 3.50GHz (8 Cores)Intel DH87RLIntel 4th Gen Core DRAM15360MB240GB OCZ VERTEX3Intel Haswell DesktopIntel Haswell HDMIVA2431Intel Connection I217-VUbuntu 13.103.10.0-999-generic (x86_64)Unity 7.0.0X Server 1.13.3intel 2.21.93.0 Mesa 9.1.3GCC 4.7.3ext41920x1080GCC 4.8.1Clang 3.2 + LLVM 3.2svnClang 3.3 + LLVM 3.3OpenBenchmarking.orgCompiler Details- GCC 4.7.3: --disable-multilib --enable-checking=release --enable-languages=c,c++,fortran- GCC 4.8.1: --disable-multilib --enable-checking=release --enable-languages=c,c++,fortran- LLVM Clang 3.2: Optimized build; Built Jun 20 2013 (14:54:23); Default target: x86_64-unknown-linux-gnu; Host CPU: x86-64 - LLVM Clang 3.3: Optimized build; Built Jun 20 2013 (12:21:18); Default target: x86_64-unknown-linux-gnu; Host CPU: x86-64 Processor Details- Scaling Governor: acpi- freq ondemand

LLVM Clang 3.3 vs. GCC 4.8 - Intel Core-AVX2 Haswellapache: Static Web Page Servingtachyon: Total Timeffmpeg: H.264 HD To NTSC DVencode-mp3: WAV To MP3smallpt: Global Illumination Renderer; 100 Samplesprimesieve: 1e12 Prime Number Generationc-ray: Total Timebuild-php: Time To Compilebuild-imagemagick: Time To Compilehimeno: Poisson Pressure Solverx264: H.264 Video Encodingtscp: AI Chess Performancescimark2: Jacobi Successive Over-Relaxationscimark2: Dense LU Matrix Factorizationscimark2: Sparse Matrix Multiplyscimark2: Fast Fourier Transformscimark2: Monte Carloblake2: Phoronix Test Suite v4.8.0m0mafft: Multiple Sequence Alignmenthmmer: Pfam Database SearchGCC 4.7.3GCC 4.8.1LLVM Clang 3.2LLVM Clang 3.325743.9912.862579.3621.4532.9293.671663.97155.336316261163.521861.551177.86251.67450.215.325.2710.4125786.1512.812579.2417.0233.3078.611593.09156.345994551164.631773.351123.80245.00615.325.305.1710.3025888.9510.4412.7712.7414027.4619.5931.941532.51155.356243231670.771774.031234.19246.79615.047.545.9611.9225295.8210.9813.1814.45142326.8527.0321.0334.351419.90153.156247491666.241827.341263.29237.86619.777.456.1910.94OpenBenchmarking.org

Apache Benchmark

Static Web Page Serving

OpenBenchmarking.orgRequests Per Second, More Is BetterApache Benchmark 2.4.3Static Web Page ServingLLVM Clang 3.2GCC 4.8.1GCC 4.7.3LLVM Clang 3.36K12K18K24K30KSE +/- 32.41, N = 3SE +/- 36.32, N = 3SE +/- 51.73, N = 3SE +/- 280.78, N = 325888.9525786.1525743.9925295.821. (CC) gcc options: -shared -fPIC -pthread -march=core-avx2 -O3

Tachyon

Total Time

OpenBenchmarking.orgSeconds, Fewer Is BetterTachyon 0.98.9Total TimeLLVM Clang 3.2LLVM Clang 3.33691215SE +/- 0.06, N = 3SE +/- 0.24, N = 610.4410.981. (CC) gcc options: -m32 -O3 -fomit-frame-pointer -ffast-math -ltachyon -lm -lpthread

FFmpeg

H.264 HD To NTSC DV

OpenBenchmarking.orgSeconds, Fewer Is BetterFFmpeg 1.1H.264 HD To NTSC DVLLVM Clang 3.2GCC 4.8.1GCC 4.7.3LLVM Clang 3.33691215SE +/- 0.04, N = 3SE +/- 0.06, N = 3SE +/- 0.08, N = 3SE +/- 0.12, N = 312.7712.8112.8613.18-Qunused-arguments-fno-tree-vectorize -MF -MT-fno-tree-vectorize -MF -MT-Qunused-arguments1. (CC) gcc options: -lavdevice -lavfilter -lavformat -lavcodec -lswresample -lswscale -lavutil -ldl -lasound -lSDL -lm -pthread -lbz2 -march=core-avx2 -O3 -std=c99 -fomit-frame-pointer -fno-math-errno -fno-signed-zeros -MMD

LAME MP3 Encoding

WAV To MP3

OpenBenchmarking.orgSeconds, Fewer Is BetterLAME MP3 Encoding 3.99.3WAV To MP3LLVM Clang 3.2LLVM Clang 3.348121620SE +/- 0.01, N = 5SE +/- 0.01, N = 512.7414.451. (CC) gcc options: -pipe -march=core-avx2 -O3 -lm

Smallpt

Global Illumination Renderer; 100 Samples

OpenBenchmarking.orgSeconds, Fewer Is BetterSmallpt 1.0Global Illumination Renderer; 100 SamplesGCC 4.7.3GCC 4.8.1LLVM Clang 3.2LLVM Clang 3.3306090120150SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 1.33, N = 325251401421. (CXX) g++ options: -fopenmp -march=core-avx2 -O3

Primesieve

1e12 Prime Number Generation

OpenBenchmarking.orgSeconds, Fewer Is BetterPrimesieve 4.21e12 Prime Number GenerationGCC 4.8.1GCC 4.7.3LLVM Clang 3.370140210280350SE +/- 0.08, N = 3SE +/- 0.06, N = 3SE +/- 2.08, N = 379.2479.36326.85-fopenmp-fopenmp1. (CXX) g++ options: -O2

C-Ray

Total Time

OpenBenchmarking.orgSeconds, Fewer Is BetterC-Ray 1.1Total TimeGCC 4.8.1GCC 4.7.3LLVM Clang 3.3LLVM Clang 3.2612182430SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.02, N = 317.0221.4527.0327.461. (CC) gcc options: -lm -lpthread -O3 -march=core-avx2

Timed PHP Compilation

Time To Compile

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed PHP Compilation 5.2.9Time To CompileLLVM Clang 3.2LLVM Clang 3.3GCC 4.7.3GCC 4.8.1816243240SE +/- 0.03, N = 3SE +/- 0.20, N = 3SE +/- 0.14, N = 3SE +/- 0.46, N = 319.5921.0332.9233.30-lpthread1. (CC) gcc options: -march=core-avx2 -O3 -pedantic -ldl -lz -lm

Timed ImageMagick Compilation

Time To Compile

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed ImageMagick Compilation 6.8.1-10Time To CompileLLVM Clang 3.2LLVM Clang 3.3GCC 4.8.1GCC 4.7.320406080100SE +/- 0.12, N = 3SE +/- 0.37, N = 3SE +/- 0.07, N = 3SE +/- 0.60, N = 331.9434.3578.6193.67

Himeno Benchmark

Poisson Pressure Solver

OpenBenchmarking.orgMFLOPS, More Is BetterHimeno Benchmark 3.0Poisson Pressure SolverGCC 4.7.3GCC 4.8.1LLVM Clang 3.2LLVM Clang 3.3400800120016002000SE +/- 0.19, N = 3SE +/- 20.45, N = 3SE +/- 2.02, N = 3SE +/- 13.80, N = 31663.971593.091532.511419.901. (CC) gcc options: -O3 -march=core-avx2

x264

H.264 Video Encoding

OpenBenchmarking.orgFrames Per Second, More Is Betterx264 2013-06-08H.264 Video EncodingGCC 4.8.1LLVM Clang 3.2GCC 4.7.3LLVM Clang 3.3306090120150SE +/- 0.40, N = 5SE +/- 0.60, N = 5SE +/- 0.33, N = 5SE +/- 0.58, N = 5156.34155.35155.33153.151. (CC) gcc options: -ldl -m64 -lm -lpthread -O3 -ffast-math -march=core-avx2 -std=gnu99 -fomit-frame-pointer -fno-tree-vectorize

TSCP

AI Chess Performance

OpenBenchmarking.orgNodes Per Second, More Is BetterTSCP 1.81AI Chess PerformanceGCC 4.7.3LLVM Clang 3.3LLVM Clang 3.2GCC 4.8.1140K280K420K560K700KSE +/- 0.00, N = 5SE +/- 480.04, N = 5SE +/- 141.40, N = 5SE +/- 520.60, N = 5631626624749624323599455

SciMark

Computational Test: Jacobi Successive Over-Relaxation

OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Jacobi Successive Over-RelaxationLLVM Clang 3.2LLVM Clang 3.3GCC 4.8.1GCC 4.7.3400800120016002000SE +/- 0.00, N = 4SE +/- 1.85, N = 4SE +/- 1.11, N = 4SE +/- 1.28, N = 41670.771666.241164.631163.52

SciMark

Computational Test: Dense LU Matrix Factorization

OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Dense LU Matrix FactorizationGCC 4.7.3LLVM Clang 3.3LLVM Clang 3.2GCC 4.8.1400800120016002000SE +/- 1.88, N = 4SE +/- 22.59, N = 4SE +/- 36.01, N = 4SE +/- 1.48, N = 41861.551827.341774.031773.35

SciMark

Computational Test: Sparse Matrix Multiply

OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Sparse Matrix MultiplyLLVM Clang 3.3LLVM Clang 3.2GCC 4.7.3GCC 4.8.130060090012001500SE +/- 5.09, N = 4SE +/- 30.06, N = 4SE +/- 0.85, N = 4SE +/- 5.13, N = 41263.291234.191177.861123.80

SciMark

Computational Test: Fast Fourier Transform

OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Fast Fourier TransformGCC 4.7.3LLVM Clang 3.2GCC 4.8.1LLVM Clang 3.350100150200250SE +/- 0.68, N = 4SE +/- 1.60, N = 4SE +/- 0.34, N = 4SE +/- 0.98, N = 4251.67246.79245.00237.86

SciMark

Computational Test: Monte Carlo

OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Monte CarloLLVM Clang 3.3GCC 4.8.1LLVM Clang 3.2GCC 4.7.3130260390520650SE +/- 0.89, N = 4SE +/- 0.00, N = 4SE +/- 5.67, N = 4SE +/- 0.55, N = 4619.77615.32615.04450.21

BLAKE2

Phoronix Test Suite v4.8.0m0

OpenBenchmarking.orgCycles Per Byte, Fewer Is BetterBLAKE2 20121223Phoronix Test Suite v4.8.0m0GCC 4.8.1GCC 4.7.3LLVM Clang 3.3LLVM Clang 3.2246810SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.13, N = 4SE +/- 0.11, N = 35.305.327.457.541. (CC) gcc options: -std=gnu99 -O3 -march=native -lcrypto -lz

Timed MAFFT Alignment

Multiple Sequence Alignment

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed MAFFT Alignment 6.864Multiple Sequence AlignmentGCC 4.8.1GCC 4.7.3LLVM Clang 3.2LLVM Clang 3.3246810SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.10, N = 3SE +/- 0.12, N = 35.175.275.966.191. (CC) gcc options: -O3 -lm -lpthread

Timed HMMer Search

Pfam Database Search

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed HMMer Search 2.3.2Pfam Database SearchGCC 4.8.1GCC 4.7.3LLVM Clang 3.3LLVM Clang 3.23691215SE +/- 0.00, N = 3SE +/- 0.02, N = 3SE +/- 0.18, N = 4SE +/- 0.00, N = 310.3010.4110.9411.921. (CC) gcc options: -march=core-avx2 -O3 -pthread -lhmmer -lsquid -lm


Phoronix Test Suite v10.8.5