POWER8 Compiler Benchmarks

Benchmarks by Michael Larabel.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 1602146-GA-POWER8COM58
Jump To Table - Results

View

Do Not Show Noisy Results
Do Not Show Results With Incomplete Data
Do Not Show Results With Little Change/Spread
List Notable Results

Limit displaying results to tests within:

C/C++ Compiler Tests 10 Tests
CPU Massive 10 Tests
Creator Workloads 3 Tests
HPC - High Performance Computing 2 Tests
Common Kernel Benchmarks 3 Tests
Multi-Core 7 Tests
Renderers 2 Tests
Server 3 Tests
Server CPU Tests 6 Tests
Single-Threaded 2 Tests
Common Workstation Benchmarks 2 Tests

Statistics

Show Overall Harmonic Mean(s)
Show Overall Geometric Mean
Show Geometric Means Per-Suite/Category
Show Wins / Losses Counts (Pie Chart)
Normalize Results
Remove Outliers Before Calculating Averages

Graph Settings

Force Line Graphs Where Applicable
Convert To Scalar Where Applicable
Prefer Vertical Bar Graphs

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Hide
Result
Result
Identifier
View Logs
Performance Per
Dollar
Date
Run
  Test
  Duration
GCC 5.3.1
February 13 2016
 
Clang 3.6.2
February 13 2016
 
Invert Hiding All Results Option
 
Only show results matching title/arguments (delimit multiple options with a comma):
Do not show results matching title/arguments (delimit multiple options with a comma):

GCC 5.3.1

Processor: POWER8 @ 3.86GHz (64 Cores), Motherboard: PowerNV 8335-GCA, Memory: 131072MB, Disk: 2000GB HDS722020ALA330, Graphics: ASPEED ASPEED Family, Network: Broadcom NetXtreme BCM5719 Gigabit PCIe

OS: Debian testing, Kernel: 4.3.0-1-powerpc64le (ppc64le), Display Driver: modesetting 1.18.0, Compiler: GCC 5.3.1 20160121 + Clang 3.6.2-3, File-System: ext4, Screen Resolution: 1024x768

Compiler Notes: --build=powerpc64le-linux-gnu --disable-browser-plugin --disable-libquadmath --disable-multilib --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-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-objc-gc --enable-plugin --enable-secureplt --enable-shared --enable-targets=powerpcle-linux --enable-threads=posix --host=powerpc64le-linux-gnu --target=powerpc64le-linux-gnu --with-arch-directory=ppc64le --with-cpu=power8 --with-default-libstdcxx-abi=new --with-long-double-128 -v
Processor Notes: Scaling Governor: powernv-cpufreq ondemand

Clang 3.6.2

OS: Debian testing, Kernel: 4.3.0-1-powerpc64le (ppc64le), Display Driver: modesetting 1.18.0, Compiler: Clang 3.6.2-3, File-System: ext4, Screen Resolution: 1024x768

Processor Notes: Scaling Governor: powernv-cpufreq ondemand

POWER8 Compiler BenchmarksOpenBenchmarking.orgPhoronix Test SuitePOWER8 @ 3.86GHz (64 Cores)PowerNV 8335-GCA131072MB2000GB HDS722020ALA330ASPEED ASPEED FamilyBroadcom NetXtreme BCM5719 Gigabit PCIeDebian testing4.3.0-1-powerpc64le (ppc64le)modesetting 1.18.0GCC 5.3.1 20160121 + Clang 3.6.2-3Clang 3.6.2-3ext41024x768ProcessorMotherboardMemoryDiskGraphicsNetworkOSKernelDisplay DriverCompilersFile-SystemScreen ResolutionPOWER8 Compiler Benchmarks PerformanceSystem Logs- GCC 5.3.1: --build=powerpc64le-linux-gnu --disable-browser-plugin --disable-libquadmath --disable-multilib --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-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-objc-gc --enable-plugin --enable-secureplt --enable-shared --enable-targets=powerpcle-linux --enable-threads=posix --host=powerpc64le-linux-gnu --target=powerpc64le-linux-gnu --with-arch-directory=ppc64le --with-cpu=power8 --with-default-libstdcxx-abi=new --with-long-double-128 -v - Scaling Governor: powernv-cpufreq ondemand

GCC 5.3.1 vs. Clang 3.6.2 ComparisonPhoronix Test SuiteBaseline+626.1%+626.1%+1252.2%+1252.2%+1878.3%+1878.3%107.2%23.8%18.2%11.2%10.3%5.8%4.5%2.8%G.I.R.1.S2504.3%Total Time136.2%J.S.O.RWAV To FLAC46.7%P.P.S30.7%CompositeTime To CompileS.M.MD.L.M.FR.4.b.P7.6%Buffer Test - Normal Load - Read Write6.9%Monte CarloF.F.TTotal TimeSmallptC-RaySciMarkFLAC Audio EncodingHimeno BenchmarkSciMarkTimed Apache CompilationSciMarkSciMarkOpenSSLPostgreSQL pgbenchSciMarkSciMarkStockfishGCC 5.3.1Clang 3.6.2

POWER8 Compiler Benchmarksrodinia: OpenMP LavaMDrodinia: OpenMP CFD Solverrodinia: OpenMP Streamclusterscimark2: Compositescimark2: Monte Carloscimark2: Fast Fourier Transformscimark2: Sparse Matrix Multiplyscimark2: Dense LU Matrix Factorizationscimark2: Jacobi Successive Over-Relaxationhimeno: Poisson Pressure Solvercompress-7zip: Compress Speed Testbuild-apache: Time To Compilec-ray: Total Timesmallpt: Global Illumination Renderer; 100 Samplesstockfish: Total Timeencode-flac: WAV To FLACopenssl: RSA 4096-bit Performancepgbench: Buffer Test - Normal Load - Read Writeapache: Static Web Page ServingGCC 5.3.1Clang 3.6.269.0157.8241.06654.58146.92189.641116.531354.68465.12347.434763032.7210.2946570030.33261.6094.6316037.33810.54155.51198.261241.751493.68963.50265.914855727.6824.311198554544.49243.1788.5416057.18OpenBenchmarking.org

Rodinia

Rodinia is a suite focused upon accelerating compute-intensive applications with accelerators. CUDA, OpenMP, and OpenCL parallel models are supported by the included applications. This profile utilizes the OpenCL and OpenMP test binaries at the moment. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 2.4Test: OpenMP LavaMDGCC 5.3.11530456075SE +/- 0.05, N = 369.011. (CXX) g++ options: -O3 -fopenmp

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 2.4Test: OpenMP CFD SolverGCC 5.3.11326395265SE +/- 0.09, N = 357.821. (CXX) g++ options: -O3 -fopenmp

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 2.4Test: OpenMP StreamclusterGCC 5.3.1918273645SE +/- 0.10, N = 341.061. (CXX) g++ options: -O3 -fopenmp

SciMark

This test runs the ANSI C version of SciMark 2.0, which is a benchmark for scientific and numerical computing developed by programmers at the National Institute of Standards and Technology. This test is made up of Fast Foruier Transform, Jacobi Successive Over-relaxation, Monte Carlo, Sparse Matrix Multiply, and dense LU matrix factorization benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: CompositeGCC 5.3.1Clang 3.6.22004006008001000SE +/- 0.33, N = 4SE +/- 1.29, N = 4654.58810.54
OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: CompositeGCC 5.3.1Clang 3.6.2140280420560700Min: 653.81 / Avg: 654.58 / Max: 655.42Min: 806.76 / Avg: 810.54 / Max: 812.52

OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Monte CarloGCC 5.3.1Clang 3.6.2306090120150SE +/- 0.07, N = 4SE +/- 0.24, N = 4146.92155.51
OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Monte CarloGCC 5.3.1Clang 3.6.2306090120150Min: 146.72 / Avg: 146.92 / Max: 147Min: 154.79 / Avg: 155.51 / Max: 155.76

OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Fast Fourier TransformGCC 5.3.1Clang 3.6.24080120160200SE +/- 0.23, N = 4SE +/- 3.21, N = 4189.64198.26
OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Fast Fourier TransformGCC 5.3.1Clang 3.6.24080120160200Min: 188.97 / Avg: 189.64 / Max: 189.97Min: 188.63 / Avg: 198.26 / Max: 201.79

OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Sparse Matrix MultiplyGCC 5.3.1Clang 3.6.230060090012001500SE +/- 1.13, N = 4SE +/- 2.23, N = 41116.531241.75
OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Sparse Matrix MultiplyGCC 5.3.1Clang 3.6.22004006008001000Min: 1114.57 / Avg: 1116.53 / Max: 1119.78Min: 1238.57 / Avg: 1241.75 / Max: 1248.37

OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Dense LU Matrix FactorizationGCC 5.3.1Clang 3.6.230060090012001500SE +/- 0.96, N = 4SE +/- 2.85, N = 41354.681493.68
OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Dense LU Matrix FactorizationGCC 5.3.1Clang 3.6.230060090012001500Min: 1351.9 / Avg: 1354.68 / Max: 1356.31Min: 1485.9 / Avg: 1493.68 / Max: 1499.58

OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Jacobi Successive Over-RelaxationGCC 5.3.1Clang 3.6.22004006008001000SE +/- 0.01, N = 4SE +/- 0.02, N = 4465.12963.50
OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Jacobi Successive Over-RelaxationGCC 5.3.1Clang 3.6.22004006008001000Min: 465.08 / Avg: 465.12 / Max: 465.14Min: 963.46 / Avg: 963.5 / Max: 963.54

Himeno Benchmark

The Himeno benchmark is a linear solver of pressure Poisson using a point-Jacobi method. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMFLOPS, More Is BetterHimeno Benchmark 3.0Poisson Pressure SolverGCC 5.3.1Clang 3.6.280160240320400SE +/- 1.77, N = 3SE +/- 0.70, N = 3347.43265.911. (CC) gcc options: -O3
OpenBenchmarking.orgMFLOPS, More Is BetterHimeno Benchmark 3.0Poisson Pressure SolverGCC 5.3.1Clang 3.6.260120180240300Min: 345.53 / Avg: 347.43 / Max: 350.96Min: 264.54 / Avg: 265.91 / Max: 266.811. (CC) gcc options: -O3

7-Zip Compression

This is a test of 7-Zip using p7zip with its integrated benchmark feature. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 9.20.1Compress Speed TestGCC 5.3.1Clang 3.6.210K20K30K40K50KSE +/- 11.89, N = 3SE +/- 60.76, N = 347630485571. (CXX) g++ options: -pipe -lpthread
OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 9.20.1Compress Speed TestGCC 5.3.1Clang 3.6.28K16K24K32K40KMin: 47607 / Avg: 47630.33 / Max: 47646Min: 48460 / Avg: 48557.33 / Max: 486691. (CXX) g++ options: -pipe -lpthread

Timed Apache Compilation

This test times how long it takes to build the Apache HTTP Server. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Apache Compilation 2.4.7Time To CompileGCC 5.3.1Clang 3.6.2816243240SE +/- 0.04, N = 3SE +/- 0.07, N = 332.7227.68
OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Apache Compilation 2.4.7Time To CompileGCC 5.3.1Clang 3.6.2714212835Min: 32.65 / Avg: 32.72 / Max: 32.79Min: 27.55 / Avg: 27.68 / Max: 27.76

C-Ray

This is a test of C-Ray, a simple raytracer designed to test the floating-point CPU performance. This test is multi-threaded (16 threads per core), will shoot 8 rays per pixel for anti-aliasing, and will generate a 1600 x 1200 image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterC-Ray 1.1Total TimeGCC 5.3.1Clang 3.6.2612182430SE +/- 0.02, N = 3SE +/- 0.02, N = 310.2924.311. (CC) gcc options: -lm -lpthread -O3
OpenBenchmarking.orgSeconds, Fewer Is BetterC-Ray 1.1Total TimeGCC 5.3.1Clang 3.6.2612182430Min: 10.26 / Avg: 10.29 / Max: 10.32Min: 24.29 / Avg: 24.31 / Max: 24.341. (CC) gcc options: -lm -lpthread -O3

Smallpt

OpenBenchmarking.orgSeconds, Fewer Is BetterSmallpt 1.0Global Illumination Renderer; 100 SamplesGCC 5.3.1Clang 3.6.230060090012001500SE +/- 0.33, N = 3SE +/- 0.00, N = 34611981. (CXX) g++ options: -fopenmp
OpenBenchmarking.orgSeconds, Fewer Is BetterSmallpt 1.0Global Illumination Renderer; 100 SamplesGCC 5.3.1Clang 3.6.22004006008001000Min: 46 / Avg: 46.33 / Max: 47Min: 1198 / Avg: 1198 / Max: 11981. (CXX) g++ options: -fopenmp

Stockfish

OpenBenchmarking.orgms, Fewer Is BetterStockfish 2014-11-26Total TimeGCC 5.3.1Clang 3.6.212002400360048006000SE +/- 1.86, N = 3SE +/- 1.15, N = 357005545-flto1. (CXX) g++ options: -lpthread -fno-exceptions -fno-rtti -ansi -pedantic -O3
OpenBenchmarking.orgms, Fewer Is BetterStockfish 2014-11-26Total TimeGCC 5.3.1Clang 3.6.210002000300040005000Min: 5696 / Avg: 5699.67 / Max: 5702Min: 5543 / Avg: 5545 / Max: 55471. (CXX) g++ options: -lpthread -fno-exceptions -fno-rtti -ansi -pedantic -O3

FLAC Audio Encoding

This test times how long it takes to encode a sample WAV file to FLAC format three times. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterFLAC Audio Encoding 1.3.1WAV To FLACGCC 5.3.1Clang 3.6.21020304050SE +/- 0.02, N = 5SE +/- 0.01, N = 530.3344.49-fvisibility=hidden1. (CXX) g++ options: -O2 -lm
OpenBenchmarking.orgSeconds, Fewer Is BetterFLAC Audio Encoding 1.3.1WAV To FLACGCC 5.3.1Clang 3.6.2918273645Min: 30.28 / Avg: 30.33 / Max: 30.38Min: 44.47 / Avg: 44.49 / Max: 44.511. (CXX) g++ options: -O2 -lm

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test measures the RSA 4096-bit performance of OpenSSL. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSigns Per Second, More Is BetterOpenSSL 1.0.1gRSA 4096-bit PerformanceGCC 5.3.1Clang 3.6.260120180240300SE +/- 0.17, N = 3SE +/- 0.15, N = 3261.60243.171. (CC) gcc options: -O3 -fomit-frame-pointer -lssl -lcrypto -ldl
OpenBenchmarking.orgSigns Per Second, More Is BetterOpenSSL 1.0.1gRSA 4096-bit PerformanceGCC 5.3.1Clang 3.6.250100150200250Min: 261.3 / Avg: 261.6 / Max: 261.9Min: 242.9 / Avg: 243.17 / Max: 243.41. (CC) gcc options: -O3 -fomit-frame-pointer -lssl -lcrypto -ldl

PostgreSQL pgbench

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 9.4.3Scaling: Buffer Test - Test: Normal Load - Mode: Read WriteGCC 5.3.1Clang 3.6.220406080100SE +/- 6.85, N = 6SE +/- 5.39, N = 694.6388.54-pthreads1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -pthread -lpgcommon -lpgport -lpq -lpthread -lrt -lcrypt -ldl -lm
OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 9.4.3Scaling: Buffer Test - Test: Normal Load - Mode: Read WriteGCC 5.3.1Clang 3.6.220406080100Min: 77.49 / Avg: 94.63 / Max: 119.16Min: 66.56 / Avg: 88.54 / Max: 101.671. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -pthread -lpgcommon -lpgport -lpq -lpthread -lrt -lcrypt -ldl -lm

Apache Benchmark

This is a test of ab, which is the Apache benchmark program. This test profile measures how many requests per second a given system can sustain when carrying out 1,000,000 requests with 100 requests being carried out concurrently. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is BetterApache Benchmark 2.4.7Static Web Page ServingGCC 5.3.1Clang 3.6.23K6K9K12K15KSE +/- 160.69, N = 3SE +/- 76.28, N = 316037.3316057.181. (CC) gcc options: -shared -fPIC -O2 -pthread
OpenBenchmarking.orgRequests Per Second, More Is BetterApache Benchmark 2.4.7Static Web Page ServingGCC 5.3.1Clang 3.6.23K6K9K12K15KMin: 15737.7 / Avg: 16037.33 / Max: 16287.79Min: 15908.75 / Avg: 16057.18 / Max: 16161.931. (CC) gcc options: -shared -fPIC -O2 -pthread

GCC 5.3.1

Processor: POWER8 @ 3.86GHz (64 Cores), Motherboard: PowerNV 8335-GCA, Memory: 131072MB, Disk: 2000GB HDS722020ALA330, Graphics: ASPEED ASPEED Family, Network: Broadcom NetXtreme BCM5719 Gigabit PCIe

OS: Debian testing, Kernel: 4.3.0-1-powerpc64le (ppc64le), Display Driver: modesetting 1.18.0, Compiler: GCC 5.3.1 20160121 + Clang 3.6.2-3, File-System: ext4, Screen Resolution: 1024x768

Compiler Notes: --build=powerpc64le-linux-gnu --disable-browser-plugin --disable-libquadmath --disable-multilib --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-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-objc-gc --enable-plugin --enable-secureplt --enable-shared --enable-targets=powerpcle-linux --enable-threads=posix --host=powerpc64le-linux-gnu --target=powerpc64le-linux-gnu --with-arch-directory=ppc64le --with-cpu=power8 --with-default-libstdcxx-abi=new --with-long-double-128 -v
Processor Notes: Scaling Governor: powernv-cpufreq ondemand

Testing initiated at 13 February 2016 20:07 by user root.

Clang 3.6.2

Processor: POWER8 @ 3.86GHz (64 Cores), Motherboard: PowerNV 8335-GCA, Memory: 131072MB, Disk: 2000GB HDS722020ALA330, Graphics: ASPEED ASPEED Family, Network: Broadcom NetXtreme BCM5719 Gigabit PCIe

OS: Debian testing, Kernel: 4.3.0-1-powerpc64le (ppc64le), Display Driver: modesetting 1.18.0, Compiler: Clang 3.6.2-3, File-System: ext4, Screen Resolution: 1024x768

Processor Notes: Scaling Governor: powernv-cpufreq ondemand

Testing initiated at 13 February 2016 22:00 by user root.