GCC9 POWER9 Compiler Benchmarks POWER9 testing with a PowerNV C1P9S01 REV 1.01 compiler benchmarking for a future article. GCC 9.1.0: Processor: POWER9 @ 3.80GHz (4 Cores / 16 Threads), Motherboard: PowerNV C1P9S01 REV 1.01, Memory: 131072MB, Disk: 1024GB SAMSUNG MZVLB1T0HALR-000L7, Graphics: ASPEED, Network: 3 x Broadcom NetXtreme BCM5719 PCIe OS: Ubuntu 19.04, Kernel: 5.0.0-17-generic (ppc64le), Compiler: GCC 9.1.0, File-System: ext4, Screen Resolution: 1024x768 GCC 10.0.0 20190616: Processor: POWER9 @ 3.80GHz (4 Cores / 16 Threads), Motherboard: PowerNV C1P9S01 REV 1.01, Memory: 131072MB, Disk: 1024GB SAMSUNG MZVLB1T0HALR-000L7, Graphics: ASPEED, Network: 3 x Broadcom NetXtreme BCM5719 PCIe OS: Ubuntu 19.04, Kernel: 5.0.0-17-generic (ppc64le), Compiler: GCC 10.0.0 20190616, File-System: ext4, Screen Resolution: 1024x768 LLVM Clang 8.0.1: Processor: POWER9 @ 3.80GHz (4 Cores / 16 Threads), Motherboard: PowerNV C1P9S01 REV 1.01, Memory: 131072MB, Disk: 1024GB SAMSUNG MZVLB1T0HALR-000L7, Graphics: ASPEED, Network: 3 x Broadcom NetXtreme BCM5719 PCIe OS: Ubuntu 19.04, Kernel: 5.0.0-17-generic (ppc64le), Compiler: Clang 8.0.1 + LLVM 8.0.1, File-System: ext4, Screen Resolution: 1024x768 SciMark 2.0 Computational Test: Composite Mflops > Higher Is Better GCC 9.1.0 ........... 268.65 |=============================================== GCC 10.0.0 20190616 . 263.02 |============================================== LLVM Clang 8.0.1 .... 277.43 |================================================= SciMark 2.0 Computational Test: Monte Carlo Mflops > Higher Is Better GCC 9.1.0 ........... 59.95 |=========================================== GCC 10.0.0 20190616 . 61.29 |============================================ LLVM Clang 8.0.1 .... 68.98 |================================================== SciMark 2.0 Computational Test: Fast Fourier Transform Mflops > Higher Is Better GCC 9.1.0 ........... 163.73 |============================================== GCC 10.0.0 20190616 . 164.05 |============================================== LLVM Clang 8.0.1 .... 174.48 |================================================= SciMark 2.0 Computational Test: Sparse Matrix Multiply Mflops > Higher Is Better GCC 9.1.0 ........... 273.69 |================================================= GCC 10.0.0 20190616 . 274.26 |================================================= LLVM Clang 8.0.1 .... 269.37 |================================================ SciMark 2.0 Computational Test: Dense LU Matrix Factorization Mflops > Higher Is Better GCC 9.1.0 ........... 283.27 |================================================= GCC 10.0.0 20190616 . 255.18 |============================================ LLVM Clang 8.0.1 .... 285.92 |================================================= SciMark 2.0 Computational Test: Jacobi Successive Over-Relaxation Mflops > Higher Is Better GCC 9.1.0 ........... 562.62 |=============================================== GCC 10.0.0 20190616 . 560.31 |=============================================== LLVM Clang 8.0.1 .... 588.43 |================================================= x264 2018-09-25 H.264 Video Encoding Frames Per Second > Higher Is Better GCC 9.1.0 ........... 15.71 |================================================== GCC 10.0.0 20190616 . 13.63 |=========================================== LLVM Clang 8.0.1 .... 15.33 |================================================= x265 3.0 H.265 1080p Video Encoding Frames Per Second > Higher Is Better GCC 9.1.0 ........... 3.92 |=================================================== GCC 10.0.0 20190616 . 3.31 |=========================================== GraphicsMagick 1.3.30 Operation: Swirl Iterations Per Minute > Higher Is Better GCC 9.1.0 ........... 131 |==================================================== GCC 10.0.0 20190616 . 128 |=================================================== LLVM Clang 8.0.1 .... 44 |================= GraphicsMagick 1.3.30 Operation: Rotate Iterations Per Minute > Higher Is Better GCC 9.1.0 ........... 185 |==================================================== GCC 10.0.0 20190616 . 184 |==================================================== LLVM Clang 8.0.1 .... 158 |============================================ GraphicsMagick 1.3.30 Operation: Sharpen Iterations Per Minute > Higher Is Better GCC 9.1.0 ........... 74 |===================================================== GCC 10.0.0 20190616 . 69 |================================================= LLVM Clang 8.0.1 .... 18 |============= GraphicsMagick 1.3.30 Operation: Enhanced Iterations Per Minute > Higher Is Better GCC 9.1.0 ........... 85 |===================================================== GCC 10.0.0 20190616 . 83 |==================================================== LLVM Clang 8.0.1 .... 22 |============== GraphicsMagick 1.3.30 Operation: Resizing Iterations Per Minute > Higher Is Better GCC 9.1.0 ........... 168 |==================================================== GCC 10.0.0 20190616 . 165 |=================================================== LLVM Clang 8.0.1 .... 85 |========================== GraphicsMagick 1.3.30 Operation: Noise-Gaussian Iterations Per Minute > Higher Is Better GCC 9.1.0 ........... 77 |===================================================== GCC 10.0.0 20190616 . 75 |==================================================== LLVM Clang 8.0.1 .... 16 |=========== GraphicsMagick 1.3.30 Operation: HWB Color Space Iterations Per Minute > Higher Is Better GCC 9.1.0 ........... 179 |==================================================== GCC 10.0.0 20190616 . 179 |==================================================== LLVM Clang 8.0.1 .... 116 |================================== Himeno Benchmark 3.0 Poisson Pressure Solver MFLOPS > Higher Is Better GCC 9.1.0 ........... 637.16 |================================================= GCC 10.0.0 20190616 . 636.79 |================================================= LLVM Clang 8.0.1 .... 581.59 |============================================= Timed LLVM Compilation 6.0.1 Time To Compile Seconds < Lower Is Better GCC 9.1.0 ........... 1260.99 |=============================================== GCC 10.0.0 20190616 . 1280.22 |================================================ LLVM Clang 8.0.1 .... 1060.28 |======================================== C-Ray 1.1 Total Time - 4K, 16 Rays Per Pixel Seconds < Lower Is Better GCC 9.1.0 ........... 204.93 |=============== GCC 10.0.0 20190616 . 205.35 |=============== LLVM Clang 8.0.1 .... 667.90 |================================================= Smallpt 1.0 Global Illumination Renderer; 128 Samples Seconds < Lower Is Better GCC 9.1.0 ........... 38.01 |================================================== GCC 10.0.0 20190616 . 37.68 |================================================== dav1d 0.3 Video Input: Summer Nature 4K Seconds < Lower Is Better GCC 9.1.0 ........... 382.25 |================================================ GCC 10.0.0 20190616 . 378.12 |=============================================== LLVM Clang 8.0.1 .... 392.45 |================================================= dav1d 0.3 Video Input: Summer Nature 1080p Seconds < Lower Is Better GCC 9.1.0 ........... 93.60 |================================================== GCC 10.0.0 20190616 . 92.02 |================================================= LLVM Clang 8.0.1 .... 90.14 |================================================ FLAC Audio Encoding 1.3.2 WAV To FLAC Seconds < Lower Is Better GCC 9.1.0 ........... 40.33 |========================================= GCC 10.0.0 20190616 . 40.32 |========================================= LLVM Clang 8.0.1 .... 48.75 |================================================== LAME MP3 Encoding 3.100 WAV To MP3 Seconds < Lower Is Better GCC 9.1.0 ........... 75.30 |================================================== GCC 10.0.0 20190616 . 75.84 |================================================== LLVM Clang 8.0.1 .... 19.29 |============= OpenSSL 1.1.1 RSA 4096-bit Performance Signs Per Second > Higher Is Better GCC 9.1.0 ........... 852.30 |================================================= GCC 10.0.0 20190616 . 838.97 |================================================ LLVM Clang 8.0.1 .... 791.23 |============================================= libjpeg-turbo tjbench 1.5.3 Test: Decompression Throughput Megapixels/sec > Higher Is Better GCC 9.1.0 ........... 101.36 |=============================================== GCC 10.0.0 20190616 . 101.14 |=============================================== LLVM Clang 8.0.1 .... 105.05 |================================================= PostgreSQL pgbench 10.3 Scaling: Buffer Test - Test: Normal Load - Mode: Read Only TPS > Higher Is Better GCC 9.1.0 ........... 96860.69 |=============================================== GCC 10.0.0 20190616 . 96945.17 |=============================================== LLVM Clang 8.0.1 .... 94535.80 |============================================== PostgreSQL pgbench 10.3 Scaling: Buffer Test - Test: Normal Load - Mode: Read Write TPS > Higher Is Better GCC 9.1.0 ........... 10960.37 |=============================================== GCC 10.0.0 20190616 . 10999.59 |=============================================== LLVM Clang 8.0.1 .... 10917.57 |=============================================== CppPerformanceBenchmarks 9 Test: Atol Seconds < Lower Is Better GCC 9.1.0 ........... 122.80 |================================================= GCC 10.0.0 20190616 . 121.04 |================================================ LLVM Clang 8.0.1 .... 120.21 |================================================ CppPerformanceBenchmarks 9 Test: Ctype Seconds < Lower Is Better GCC 9.1.0 ........... 167.77 |================================================= GCC 10.0.0 20190616 . 167.57 |================================================= LLVM Clang 8.0.1 .... 153.26 |============================================= CppPerformanceBenchmarks 9 Test: Math Library Seconds < Lower Is Better GCC 9.1.0 ........... 781.59 |================================================= GCC 10.0.0 20190616 . 780.84 |================================================= LLVM Clang 8.0.1 .... 775.78 |================================================= CppPerformanceBenchmarks 9 Test: Random Numbers Seconds < Lower Is Better GCC 9.1.0 ........... 2156.82 |======================= GCC 10.0.0 20190616 . 2137.23 |======================= LLVM Clang 8.0.1 .... 4529.86 |================================================ CppPerformanceBenchmarks 9 Test: Stepanov Vector Seconds < Lower Is Better GCC 9.1.0 ........... 139.65 |================================================= GCC 10.0.0 20190616 . 139.74 |================================================= LLVM Clang 8.0.1 .... 131.37 |============================================== CppPerformanceBenchmarks 9 Test: Function Objects Seconds < Lower Is Better GCC 9.1.0 ........... 26.41 |=============================================== GCC 10.0.0 20190616 . 26.39 |=============================================== LLVM Clang 8.0.1 .... 27.88 |================================================== CppPerformanceBenchmarks 9 Test: Stepanov Abstraction Seconds < Lower Is Better GCC 9.1.0 ........... 57.19 |================================================== GCC 10.0.0 20190616 . 57.18 |================================================== LLVM Clang 8.0.1 .... 53.07 |============================================== Redis 4.0.8 Test: LPOP Requests Per Second > Higher Is Better GCC 9.1.0 ........... 1059020.41 |============================================= GCC 10.0.0 20190616 . 1045275.77 |============================================ LLVM Clang 8.0.1 .... 1007820.41 |=========================================== Redis 4.0.8 Test: SADD Requests Per Second > Higher Is Better GCC 9.1.0 ........... 710532.17 |============================================= GCC 10.0.0 20190616 . 725169.62 |============================================== LLVM Clang 8.0.1 .... 728590.02 |============================================== Redis 4.0.8 Test: LPUSH Requests Per Second > Higher Is Better GCC 9.1.0 ........... 568842.41 |============================================== GCC 10.0.0 20190616 . 574786.06 |============================================== LLVM Clang 8.0.1 .... 571604.42 |============================================== Redis 4.0.8 Test: GET Requests Per Second > Higher Is Better GCC 9.1.0 ........... 934311.06 |=========================================== GCC 10.0.0 20190616 . 942834.62 |============================================ LLVM Clang 8.0.1 .... 993911.90 |============================================== Redis 4.0.8 Test: SET Requests Per Second > Higher Is Better GCC 9.1.0 ........... 585363.53 |============================================== GCC 10.0.0 20190616 . 571452.36 |============================================= LLVM Clang 8.0.1 .... 571064.36 |============================================= Apache Benchmark 2.4.29 Static Web Page Serving Requests Per Second > Higher Is Better GCC 9.1.0 ........... 17985.14 |=============================================== GCC 10.0.0 20190616 . 17853.33 |=============================================== LLVM Clang 8.0.1 .... 17819.95 |===============================================