Ryzen 9 3900X Znver2 Compiler Tuning AMD Ryzen 9 3900X 12-Core testing of GCC 9 and GCC 10 development with Znver2 tuning following recent cost table updates, etc. Benchmarks by Michael Larabel for a future article.. GCC 9.1.0: Processor: AMD Ryzen 9 3900X 12-Core @ 3.80GHz (12 Cores / 24 Threads), Motherboard: ASUS ROG CROSSHAIR VIII HERO (WI-FI) (0702 BIOS), Chipset: AMD Device 1480, Memory: 16384MB, Disk: 2000GB Force MP600, Graphics: Sapphire AMD Radeon RX 550 640SP / 560/560X 4GB (1300/1750MHz), Audio: AMD Device aae0, Monitor: ASUS VP28U, Network: Realtek Device 8125 + Intel I211 + Intel Device 2723 OS: Ubuntu 18.04, Kernel: 5.3.0-999-generic (x86_64) 20190725, Desktop: GNOME Shell 3.28.4, Display Server: X Server 1.20.4, Display Driver: modesetting 1.20.4, OpenGL: 4.5 Mesa 19.0.2 (LLVM 8.0.0), Compiler: GCC 9.1.0, File-System: ext4, Screen Resolution: 3840x2160 GCC 9.1.0 znver2: Processor: AMD Ryzen 9 3900X 12-Core @ 3.80GHz (12 Cores / 24 Threads), Motherboard: ASUS ROG CROSSHAIR VIII HERO (WI-FI) (0702 BIOS), Chipset: AMD Device 1480, Memory: 16384MB, Disk: 2000GB Force MP600, Graphics: Sapphire AMD Radeon RX 550 640SP / 560/560X 4GB (1300/1750MHz), Audio: AMD Device aae0, Monitor: ASUS VP28U, Network: Realtek Device 8125 + Intel I211 + Intel Device 2723 OS: Ubuntu 18.04, Kernel: 5.3.0-999-generic (x86_64) 20190725, Desktop: GNOME Shell 3.28.4, Display Server: X Server 1.20.4, Display Driver: modesetting 1.20.4, OpenGL: 4.5 Mesa 19.0.2 (LLVM 8.0.0), Compiler: GCC 9.1.0, File-System: ext4, Screen Resolution: 3840x2160 GCC 10.0.0 znver2: Processor: AMD Ryzen 9 3900X 12-Core @ 3.80GHz (12 Cores / 24 Threads), Motherboard: ASUS ROG CROSSHAIR VIII HERO (WI-FI) (0702 BIOS), Chipset: AMD Device 1480, Memory: 16384MB, Disk: 2000GB Force MP600, Graphics: Sapphire AMD Radeon RX 550 640SP / 560/560X 4GB (1300/1750MHz), Audio: AMD Device aae0, Monitor: ASUS VP28U, Network: Realtek Device 8125 + Intel I211 + Intel Device 2723 OS: Ubuntu 18.04, Kernel: 5.3.0-999-generic (x86_64) 20190725, Desktop: GNOME Shell 3.28.4, Display Server: X Server 1.20.4, Display Driver: modesetting 1.20.4, OpenGL: 4.5 Mesa 19.0.2 (LLVM 8.0.0), Compiler: GCC 10.0.0 20190727, File-System: ext4, Screen Resolution: 3840x2160 GCC 10.0.0: Processor: AMD Ryzen 9 3900X 12-Core @ 3.80GHz (12 Cores / 24 Threads), Motherboard: ASUS ROG CROSSHAIR VIII HERO (WI-FI) (0702 BIOS), Chipset: AMD Device 1480, Memory: 16384MB, Disk: 2000GB Force MP600, Graphics: Sapphire AMD Radeon RX 550 640SP / 560/560X 4GB (1300/1750MHz), Audio: AMD Device aae0, Monitor: ASUS VP28U, Network: Realtek Device 8125 + Intel I211 + Intel Device 2723 OS: Ubuntu 18.04, Kernel: 5.3.0-999-generic (x86_64) 20190725, Desktop: GNOME Shell 3.28.4, Display Server: X Server 1.20.4, Display Driver: modesetting 1.20.4, OpenGL: 4.5 Mesa 19.0.2 (LLVM 8.0.0), Compiler: GCC 10.0.0 20190727, File-System: ext4, Screen Resolution: 3840x2160 AOBench Size: 2048 x 2048 - Total Time Seconds < Lower Is Better GCC 9.1.0 ......... 34.60 |================================================== GCC 9.1.0 znver2 .. 33.20 |================================================ GCC 10.0.0 znver2 . 33.05 |================================================ GCC 10.0.0 ........ 35.98 |==================================================== AOM AV1 2019-02-11 AV1 Video Encoding Frames Per Second > Higher Is Better GCC 9.1.0 ......... 0.27 |============================================= GCC 9.1.0 znver2 .. 0.31 |=================================================== GCC 10.0.0 znver2 . 0.32 |===================================================== GCC 10.0.0 ........ 0.27 |============================================= Apache Benchmark 2.4.29 Static Web Page Serving Requests Per Second > Higher Is Better GCC 9.1.0 ......... 38392.29 |================================================= GCC 9.1.0 znver2 .. 38022.79 |================================================ GCC 10.0.0 znver2 . 38009.25 |================================================ GCC 10.0.0 ........ 38490.98 |================================================= Apache Siege 2.4.29 Concurrent Users: 200 Transactions Per Second > Higher Is Better GCC 9.1.0 ......... 60835.79 |============================== GCC 9.1.0 znver2 .. 99824.49 |================================================= GCC 10.0.0 znver2 . 83275.06 |========================================= GCC 10.0.0 ........ 82293.14 |======================================== Apache Siege 2.4.29 Concurrent Users: 250 Transactions Per Second > Higher Is Better GCC 9.1.0 ......... 98050.91 |============================================== GCC 9.1.0 znver2 .. 96842.13 |============================================= GCC 10.0.0 znver2 . 102423.07 |================================================ GCC 10.0.0 ........ 62725.24 |============================= Bullet Physics Engine 2.81 Test: Raytests Seconds < Lower Is Better GCC 9.1.0 ......... 1.98 |================================================== GCC 9.1.0 znver2 .. 2.04 |=================================================== GCC 10.0.0 znver2 . 2.06 |==================================================== GCC 10.0.0 ........ 2.11 |===================================================== Bullet Physics Engine 2.81 Test: 3000 Fall Seconds < Lower Is Better GCC 9.1.0 ......... 3.20 |================================================= GCC 9.1.0 znver2 .. 3.22 |================================================== GCC 10.0.0 znver2 . 3.27 |================================================== GCC 10.0.0 ........ 3.44 |===================================================== Bullet Physics Engine 2.81 Test: 1000 Stack Seconds < Lower Is Better GCC 9.1.0 ......... 3.88 |================================================== GCC 9.1.0 znver2 .. 3.77 |================================================= GCC 10.0.0 znver2 . 3.85 |================================================== GCC 10.0.0 ........ 4.11 |===================================================== Bullet Physics Engine 2.81 Test: 1000 Convex Seconds < Lower Is Better GCC 9.1.0 ......... 3.51 |================================================== GCC 9.1.0 znver2 .. 3.57 |=================================================== GCC 10.0.0 znver2 . 3.60 |=================================================== GCC 10.0.0 ........ 3.73 |===================================================== Bullet Physics Engine 2.81 Test: 136 Ragdolls Seconds < Lower Is Better GCC 9.1.0 ......... 2.05 |================================================= GCC 9.1.0 znver2 .. 2.04 |================================================= GCC 10.0.0 znver2 . 2.05 |================================================= GCC 10.0.0 ........ 2.20 |===================================================== Bullet Physics Engine 2.81 Test: Prim Trimesh Seconds < Lower Is Better GCC 9.1.0 ......... 0.75 |================================================= GCC 9.1.0 znver2 .. 0.77 |================================================== GCC 10.0.0 znver2 . 0.77 |================================================== GCC 10.0.0 ........ 0.81 |===================================================== Bullet Physics Engine 2.81 Test: Convex Trimesh Seconds < Lower Is Better GCC 9.1.0 ......... 0.89 |================================================== GCC 9.1.0 znver2 .. 0.90 |================================================== GCC 10.0.0 znver2 . 0.91 |=================================================== GCC 10.0.0 ........ 0.95 |===================================================== C-Ray 1.1 Total Time - 4K, 16 Rays Per Pixel Seconds < Lower Is Better GCC 9.1.0 ......... 43.09 |==================================================== GCC 9.1.0 znver2 .. 39.42 |================================================ GCC 10.0.0 znver2 . 39.36 |=============================================== GCC 10.0.0 ........ 42.63 |=================================================== Coremark 1.0 CoreMark Size 666 - Iterations Per Second Iterations/Sec > Higher Is Better GCC 9.1.0 ......... 567987.34 |================================================ GCC 9.1.0 znver2 .. 555154.60 |=============================================== GCC 10.0.0 znver2 . 567096.65 |================================================ GCC 10.0.0 ........ 568329.00 |================================================ CppPerformanceBenchmarks 9 Test: Atol Seconds < Lower Is Better GCC 9.1.0 ......... 59.31 |================================================= GCC 9.1.0 znver2 .. 60.38 |================================================== GCC 10.0.0 znver2 . 63.34 |==================================================== GCC 10.0.0 ........ 59.97 |================================================= CppPerformanceBenchmarks 9 Test: Ctype Seconds < Lower Is Better GCC 9.1.0 ......... 31.52 |==================================================== GCC 9.1.0 znver2 .. 31.43 |==================================================== GCC 10.0.0 znver2 . 31.30 |==================================================== GCC 10.0.0 ........ 31.51 |==================================================== CppPerformanceBenchmarks 9 Test: Math Library Seconds < Lower Is Better GCC 9.1.0 ......... 309.36 |=================================================== GCC 9.1.0 znver2 .. 302.82 |================================================== GCC 10.0.0 znver2 . 306.02 |================================================== GCC 10.0.0 ........ 307.23 |=================================================== CppPerformanceBenchmarks 9 Test: Random Numbers Seconds < Lower Is Better GCC 9.1.0 ......... 751.15 |================================================ GCC 9.1.0 znver2 .. 750.66 |================================================ GCC 10.0.0 znver2 . 787.77 |================================================== GCC 10.0.0 ........ 799.88 |=================================================== CppPerformanceBenchmarks 9 Test: Stepanov Vector Seconds < Lower Is Better GCC 9.1.0 ......... 76.45 |=================================================== GCC 9.1.0 znver2 .. 74.08 |================================================== GCC 10.0.0 znver2 . 77.22 |==================================================== GCC 10.0.0 ........ 74.26 |================================================== CppPerformanceBenchmarks 9 Test: Function Objects Seconds < Lower Is Better GCC 9.1.0 ......... 14.40 |================================================== GCC 9.1.0 znver2 .. 14.15 |================================================= GCC 10.0.0 znver2 . 14.90 |=================================================== GCC 10.0.0 ........ 15.10 |==================================================== CppPerformanceBenchmarks 9 Test: Stepanov Abstraction Seconds < Lower Is Better GCC 9.1.0 ......... 27.60 |================================================== GCC 9.1.0 znver2 .. 28.93 |==================================================== GCC 10.0.0 znver2 . 28.30 |=================================================== GCC 10.0.0 ........ 28.19 |=================================================== Cpuminer-Opt 3.8.8.1 Algorithm: m7m kH/s - Hash Speed > Higher Is Better GCC 9.1.0 ......... 593.89 |=================================================== GCC 9.1.0 znver2 .. 590.66 |=================================================== GCC 10.0.0 znver2 . 590.80 |=================================================== GCC 10.0.0 ........ 591.32 |=================================================== Cpuminer-Opt 3.8.8.1 Algorithm: deep kH/s - Hash Speed > Higher Is Better GCC 9.1.0 ......... 11190.00 |================================================= GCC 9.1.0 znver2 .. 10230.34 |============================================= GCC 10.0.0 znver2 . 11123.00 |================================================= GCC 10.0.0 ........ 11137.00 |================================================= Cpuminer-Opt 3.8.8.1 Algorithm: lbry kH/s - Hash Speed > Higher Is Better GCC 9.1.0 ......... 34583 |=================================================== GCC 9.1.0 znver2 .. 34420 |=================================================== GCC 10.0.0 znver2 . 34630 |=================================================== GCC 10.0.0 ........ 35288 |==================================================== Cpuminer-Opt 3.8.8.1 Algorithm: skein kH/s - Hash Speed > Higher Is Better GCC 9.1.0 ......... 39397 |=================================================== GCC 9.1.0 znver2 .. 39797 |==================================================== GCC 10.0.0 znver2 . 39843 |==================================================== GCC 10.0.0 ........ 39720 |==================================================== Cpuminer-Opt 3.8.8.1 Algorithm: myr-gr kH/s - Hash Speed > Higher Is Better GCC 9.1.0 ......... 14127 |==================================================== GCC 9.1.0 znver2 .. 14023 |==================================================== GCC 10.0.0 znver2 . 14137 |==================================================== GCC 10.0.0 ........ 14130 |==================================================== Cpuminer-Opt 3.8.8.1 Algorithm: sha256t kH/s - Hash Speed > Higher Is Better GCC 9.1.0 ......... 87951 |==================================================== GCC 9.1.0 znver2 .. 87238 |==================================================== GCC 10.0.0 znver2 . 86440 |=================================================== GCC 10.0.0 ........ 86417 |=================================================== FFmpeg 4.0.2 H.264 HD To NTSC DV Seconds < Lower Is Better GCC 9.1.0 ......... 6.86 |===================================================== GCC 9.1.0 znver2 .. 6.83 |===================================================== GCC 10.0.0 znver2 . 6.78 |==================================================== GCC 10.0.0 ........ 6.88 |===================================================== FFTW 3.3.6 Build: Stock - Size: 1D FFT Size 32 Mflops > Higher Is Better GCC 9.1.0 ......... 12958 |================================================ GCC 9.1.0 znver2 .. 11828 |============================================ GCC 10.0.0 znver2 . 14113 |==================================================== GCC 10.0.0 ........ 12748 |=============================================== FFTW 3.3.6 Build: Stock - Size: 2D FFT Size 32 Mflops > Higher Is Better GCC 9.1.0 ......... 11909 |=========================================== GCC 9.1.0 znver2 .. 14314 |==================================================== GCC 10.0.0 znver2 . 14119 |=================================================== GCC 10.0.0 ........ 12902 |=============================================== FFTW 3.3.6 Build: Stock - Size: 2D FFT Size 512 Mflops > Higher Is Better GCC 9.1.0 ......... 9028.10 |========================================= GCC 9.1.0 znver2 .. 10814.00 |================================================= GCC 10.0.0 znver2 . 10531.00 |================================================ GCC 10.0.0 ........ 9583.73 |=========================================== FFTW 3.3.6 Build: Stock - Size: 2D FFT Size 4096 Mflops > Higher Is Better GCC 9.1.0 ......... 7063.03 |============================================= GCC 9.1.0 znver2 .. 7920.17 |================================================== GCC 10.0.0 znver2 . 7823.27 |================================================= GCC 10.0.0 ........ 7071.30 |============================================= FFTW 3.3.6 Build: Float + SSE - Size: 2D FFT Size 32 Mflops > Higher Is Better GCC 9.1.0 ......... 45253 |=================================================== GCC 9.1.0 znver2 .. 44951 |================================================== GCC 10.0.0 znver2 . 46305 |==================================================== GCC 10.0.0 ........ 45361 |=================================================== FLAC Audio Encoding 1.3.2 WAV To FLAC Seconds < Lower Is Better GCC 9.1.0 ......... 7.70 |================================================== GCC 9.1.0 znver2 .. 7.99 |==================================================== GCC 10.0.0 znver2 . 8.11 |===================================================== GCC 10.0.0 ........ 7.72 |================================================== GNU MPC 1.1.0 Multi-Precision Benchmark Global Score > Higher Is Better GCC 9.1.0 ......... 9597 |===================================================== GCC 9.1.0 znver2 .. 9577 |===================================================== GCC 10.0.0 znver2 . 9357 |==================================================== GCC 10.0.0 ........ 9580 |===================================================== GraphicsMagick 1.3.30 Operation: Swirl Iterations Per Minute > Higher Is Better GCC 9.1.0 ......... 251 |=================================================== GCC 9.1.0 znver2 .. 259 |===================================================== GCC 10.0.0 znver2 . 264 |====================================================== GCC 10.0.0 ........ 254 |==================================================== GraphicsMagick 1.3.30 Operation: Rotate Iterations Per Minute > Higher Is Better GCC 9.1.0 ......... 262 |=================================================== GCC 9.1.0 znver2 .. 263 |=================================================== GCC 10.0.0 znver2 . 277 |====================================================== GCC 10.0.0 ........ 262 |=================================================== GraphicsMagick 1.3.30 Operation: Sharpen Iterations Per Minute > Higher Is Better GCC 9.1.0 ......... 181 |================================================== GCC 9.1.0 znver2 .. 195 |====================================================== GCC 10.0.0 znver2 . 196 |====================================================== GCC 10.0.0 ........ 181 |================================================== GraphicsMagick 1.3.30 Operation: Enhanced Iterations Per Minute > Higher Is Better GCC 9.1.0 ......... 209 |=================================================== GCC 9.1.0 znver2 .. 221 |====================================================== GCC 10.0.0 znver2 . 223 |====================================================== GCC 10.0.0 ........ 208 |================================================== GraphicsMagick 1.3.30 Operation: Resizing Iterations Per Minute > Higher Is Better GCC 9.1.0 ......... 275 |==================================================== GCC 9.1.0 znver2 .. 280 |===================================================== GCC 10.0.0 znver2 . 286 |====================================================== GCC 10.0.0 ........ 274 |==================================================== GraphicsMagick 1.3.30 Operation: Noise-Gaussian Iterations Per Minute > Higher Is Better GCC 9.1.0 ......... 170 |===================================================== GCC 9.1.0 znver2 .. 171 |===================================================== GCC 10.0.0 znver2 . 173 |====================================================== GCC 10.0.0 ........ 170 |===================================================== GraphicsMagick 1.3.30 Operation: HWB Color Space Iterations Per Minute > Higher Is Better GCC 9.1.0 ......... 287 |=================================================== GCC 9.1.0 znver2 .. 293 |==================================================== GCC 10.0.0 znver2 . 302 |====================================================== GCC 10.0.0 ........ 288 |=================================================== GROMACS 2018.3 Water Benchmark Ns Per Day > Higher Is Better GCC 9.1.0 ......... 0.98 |==================================================== GCC 9.1.0 znver2 .. 0.99 |===================================================== GCC 10.0.0 znver2 . 0.98 |==================================================== GCC 10.0.0 ........ 0.97 |==================================================== High Performance Conjugate Gradient 3.0 GFLOP/s > Higher Is Better GCC 9.1.0 ......... 1.09 |===================================================== GCC 9.1.0 znver2 .. 1.08 |===================================================== GCC 10.0.0 znver2 . 1.08 |===================================================== GCC 10.0.0 ........ 1.09 |===================================================== Himeno Benchmark 3.0 Poisson Pressure Solver MFLOPS > Higher Is Better GCC 9.1.0 ......... 1322.90 |================================================ GCC 9.1.0 znver2 .. 1378.46 |================================================== GCC 10.0.0 znver2 . 1385.88 |================================================== GCC 10.0.0 ........ 1385.23 |================================================== HPC Challenge 1.5.0 Test / Class: G-HPL GFLOPS > Higher Is Better GCC 9.1.0 ......... 70.97 |=================================================== GCC 9.1.0 znver2 .. 71.78 |==================================================== GCC 10.0.0 znver2 . 71.05 |=================================================== GCC 10.0.0 ........ 71.07 |=================================================== HPC Challenge 1.5.0 Test / Class: G-Ffte GFLOPS > Higher Is Better GCC 9.1.0 ......... 8.59803 |================================================= GCC 9.1.0 znver2 .. 8.60514 |================================================= GCC 10.0.0 znver2 . 8.63794 |================================================= GCC 10.0.0 ........ 8.81748 |================================================== HPC Challenge 1.5.0 Test / Class: G-Ffte GFLOP/s > Higher Is Better GCC 9.1.0 ......... 8.59803 |================================================= GCC 9.1.0 znver2 .. 8.60514 |================================================= GCC 10.0.0 znver2 . 8.63794 |================================================= GCC 10.0.0 ........ 8.81748 |================================================== HPC Challenge 1.5.0 Test / Class: EP-DGEMM GFLOPS > Higher Is Better GCC 9.1.0 ......... 32.83 |==================================================== GCC 9.1.0 znver2 .. 32.60 |==================================================== GCC 10.0.0 znver2 . 32.84 |==================================================== GCC 10.0.0 ........ 32.86 |==================================================== HPC Challenge 1.5.0 Test / Class: G-Ptrans GB/s > Higher Is Better GCC 9.1.0 ......... 2.73255 |============================================== GCC 9.1.0 znver2 .. 2.95225 |================================================== GCC 10.0.0 znver2 . 2.94730 |================================================== GCC 10.0.0 ........ 2.72974 |============================================== HPC Challenge 1.5.0 Test / Class: EP-STREAM Triad GB/s > Higher Is Better GCC 9.1.0 ......... 1.70820 |================================================= GCC 9.1.0 znver2 .. 1.71668 |================================================== GCC 10.0.0 znver2 . 1.73055 |================================================== GCC 10.0.0 ........ 1.72205 |================================================== HPC Challenge 1.5.0 Test / Class: G-Random Access GUP/s > Higher Is Better GCC 9.1.0 ......... 0.09757 |================================================== GCC 9.1.0 znver2 .. 0.09798 |================================================== GCC 10.0.0 znver2 . 0.09771 |================================================== GCC 10.0.0 ........ 0.09778 |================================================== HPC Challenge 1.5.0 Test / Class: Random Ring Latency usecs < Lower Is Better GCC 9.1.0 ......... 0.32596 |================================================= GCC 9.1.0 znver2 .. 0.32698 |================================================= GCC 10.0.0 znver2 . 0.32521 |================================================= GCC 10.0.0 ........ 0.33186 |================================================== HPC Challenge 1.5.0 Test / Class: Random Ring Bandwidth GB/s > Higher Is Better GCC 9.1.0 ......... 4.89161 |================================================ GCC 9.1.0 znver2 .. 4.98832 |================================================= GCC 10.0.0 znver2 . 5.04603 |================================================== GCC 10.0.0 ........ 4.94947 |================================================= HPC Challenge 1.5.0 Test / Class: Max Ping Pong Bandwidth MB/s > Higher Is Better GCC 9.1.0 ......... 24227.25 |================================================= GCC 9.1.0 znver2 .. 23832.61 |================================================ GCC 10.0.0 znver2 . 23993.04 |================================================= GCC 10.0.0 ........ 23885.44 |================================================ John The Ripper 1.9.0-jumbo-1 Test: Blowfish Real C/S > Higher Is Better GCC 9.1.0 ......... 20335 |==================================================== GCC 9.1.0 znver2 .. 20253 |==================================================== GCC 10.0.0 znver2 . 20426 |==================================================== GCC 10.0.0 ........ 20426 |==================================================== LAME MP3 Encoding 3.100 WAV To MP3 Seconds < Lower Is Better GCC 9.1.0 ......... 7.25 |==================================================== GCC 9.1.0 znver2 .. 6.94 |================================================= GCC 10.0.0 znver2 . 7.45 |===================================================== GCC 10.0.0 ........ 7.28 |==================================================== libjpeg-turbo tjbench 2.0.2 Test: Decompression Throughput Megapixels/sec > Higher Is Better GCC 9.1.0 ......... 218.09 |================================================= GCC 9.1.0 znver2 .. 225.64 |=================================================== GCC 10.0.0 znver2 . 225.44 |=================================================== GCC 10.0.0 ........ 220.33 |================================================== lzbench 2017-08-08 Test: XZ 0 - Process: Compression MB/s > Higher Is Better GCC 9.1.0 ......... 39 |====================================================== GCC 9.1.0 znver2 .. 40 |======================================================= GCC 10.0.0 znver2 . 37 |=================================================== GCC 10.0.0 ........ 40 |======================================================= lzbench 2017-08-08 Test: XZ 0 - Process: Decompression MB/s > Higher Is Better GCC 9.1.0 ......... 116 |====================================================== GCC 9.1.0 znver2 .. 116 |====================================================== GCC 10.0.0 znver2 . 108 |================================================== GCC 10.0.0 ........ 113 |===================================================== lzbench 2017-08-08 Test: Zstd 1 - Process: Compression MB/s > Higher Is Better GCC 9.1.0 ......... 468 |====================================================== GCC 9.1.0 znver2 .. 468 |====================================================== GCC 10.0.0 znver2 . 453 |==================================================== GCC 10.0.0 ........ 467 |====================================================== lzbench 2017-08-08 Test: Zstd 1 - Process: Decompression MB/s > Higher Is Better GCC 9.1.0 ......... 1269 |==================================================== GCC 9.1.0 znver2 .. 1268 |==================================================== GCC 10.0.0 znver2 . 1250 |=================================================== GCC 10.0.0 ........ 1287 |===================================================== lzbench 2017-08-08 Test: Brotli 0 - Process: Compression MB/s > Higher Is Better GCC 10.0.0 ........ 507 |==================================================== GCC 10.0.0 znver2 . 499 |==================================================== GCC 9.1.0 ......... 494 |=================================================== GCC 9.1.0 znver2 .. 515 |===================================================== GCC 10.0.0 znver2 . 505 |==================================================== GCC 10.0.0 ........ 522 |====================================================== lzbench 2017-08-08 Test: Libdeflate 1 - Process: Compression MB/s > Higher Is Better GCC 9.1.0 ......... 239 |================================================== GCC 9.1.0 znver2 .. 257 |====================================================== GCC 10.0.0 znver2 . 248 |==================================================== GCC 10.0.0 ........ 250 |===================================================== lzbench 2017-08-08 Test: Libdeflate 1 - Process: Decompression MB/s > Higher Is Better GCC 9.1.0 ......... 1119 |================================================== GCC 9.1.0 znver2 .. 1183 |===================================================== GCC 10.0.0 znver2 . 1159 |==================================================== GCC 10.0.0 ........ 1147 |=================================================== m-queens 1.2 Time To Solve Seconds < Lower Is Better GCC 9.1.0 ......... 47.12 |==================================================== GCC 9.1.0 znver2 .. 47.27 |==================================================== GCC 10.0.0 znver2 . 47.21 |==================================================== GCC 10.0.0 ........ 47.14 |==================================================== Memcached mcperf 1.5.10 Method: Get Operations Per Second > Higher Is Better GCC 9.1.0 ......... 92376.40 |=============================================== GCC 9.1.0 znver2 .. 93850.59 |=============================================== GCC 10.0.0 znver2 . 97228.27 |================================================= GCC 10.0.0 ........ 95710.60 |================================================ Memcached mcperf 1.5.10 Method: Set Operations Per Second > Higher Is Better GCC 9.1.0 ......... 52914.10 |============================================ GCC 9.1.0 znver2 .. 59232.07 |================================================= GCC 10.0.0 znver2 . 52910.87 |============================================ GCC 10.0.0 ........ 57193.25 |=============================================== MKL-DNN 2019-04-16 Harness: IP Batch 1D - Data Type: f32 ms < Lower Is Better GCC 9.1.0 ......... 152.51 |================================================= GCC 9.1.0 znver2 .. 155.34 |================================================== GCC 10.0.0 znver2 . 157.72 |=================================================== GCC 10.0.0 ........ 154.09 |================================================== MKL-DNN 2019-04-16 Harness: IP Batch All - Data Type: f32 ms < Lower Is Better GCC 9.1.0 ......... 1523.52 |================================================ GCC 9.1.0 znver2 .. 1599.68 |================================================== GCC 10.0.0 znver2 . 1556.91 |================================================= GCC 10.0.0 ........ 1582.78 |================================================= MKL-DNN 2019-04-16 Harness: Convolution Batch conv_3d - Data Type: f32 ms < Lower Is Better GCC 9.1.0 ......... 117.60 |=================================================== GCC 9.1.0 znver2 .. 118.47 |=================================================== GCC 10.0.0 znver2 . 118.02 |=================================================== GCC 10.0.0 ........ 116.62 |================================================== MKL-DNN 2019-04-16 Harness: Convolution Batch conv_all - Data Type: f32 ms < Lower Is Better GCC 9.1.0 ......... 19613.70 |================================================= GCC 9.1.0 znver2 .. 19696.57 |================================================= GCC 10.0.0 znver2 . 19694.33 |================================================= GCC 10.0.0 ........ 19803.57 |================================================= MKL-DNN 2019-04-16 Harness: Deconvolution Batch deconv_1d - Data Type: f32 ms < Lower Is Better GCC 9.1.0 ......... 221.23 |=================================================== GCC 9.1.0 znver2 .. 217.02 |================================================== GCC 10.0.0 znver2 . 218.66 |================================================== GCC 10.0.0 ........ 212.83 |================================================= MKL-DNN 2019-04-16 Harness: Deconvolution Batch deconv_3d - Data Type: f32 ms < Lower Is Better GCC 9.1.0 ......... 59.00 |==================================================== GCC 9.1.0 znver2 .. 57.97 |=================================================== GCC 10.0.0 znver2 . 56.87 |================================================== GCC 10.0.0 ........ 58.16 |=================================================== MKL-DNN 2019-04-16 Harness: Convolution Batch conv_alexnet - Data Type: f32 ms < Lower Is Better GCC 9.1.0 ......... 2527.50 |================================================== GCC 9.1.0 znver2 .. 2520.01 |================================================== GCC 10.0.0 znver2 . 2543.93 |================================================== GCC 10.0.0 ........ 2507.16 |================================================= MKL-DNN 2019-04-16 Harness: Deconvolution Batch deconv_all - Data Type: f32 ms < Lower Is Better GCC 9.1.0 ......... 51813.05 |================================================= GCC 9.1.0 znver2 .. 52238.80 |================================================= GCC 10.0.0 znver2 . 50679.53 |================================================ GCC 10.0.0 ........ 50039.13 |=============================================== MKL-DNN 2019-04-16 Harness: Convolution Batch conv_googlenet_v3 - Data Type: f32 ms < Lower Is Better GCC 9.1.0 ......... 1147.62 |================================================== GCC 9.1.0 znver2 .. 1153.46 |================================================== GCC 10.0.0 znver2 . 1145.95 |================================================== GCC 10.0.0 ........ 1145.01 |================================================== NGINX Benchmark 1.9.9 Static Web Page Serving Requests Per Second > Higher Is Better GCC 9.1.0 ......... 39734.85 |================================================= GCC 9.1.0 znver2 .. 39602.49 |================================================= GCC 10.0.0 znver2 . 39346.91 |================================================= GCC 10.0.0 ........ 39525.70 |================================================= Ogg Encoding 1.3.3 WAV To Ogg Seconds < Lower Is Better GCC 9.1.0 ......... 5.13 |=================================================== GCC 9.1.0 znver2 .. 5.05 |================================================== GCC 10.0.0 znver2 . 5.36 |===================================================== GCC 10.0.0 ........ 5.05 |================================================== OpenSSL 1.1.1 RSA 4096-bit Performance Signs Per Second > Higher Is Better GCC 9.1.0 ......... 3516.27 |================================================== GCC 9.1.0 znver2 .. 3481.50 |================================================== GCC 10.0.0 znver2 . 3487.10 |================================================== GCC 10.0.0 ........ 3492.53 |================================================== PostgreSQL pgbench 10.3 Scaling: Buffer Test - Test: Normal Load - Mode: Read Only TPS > Higher Is Better GCC 9.1.0 ......... 300353.09 |================================================ GCC 9.1.0 znver2 .. 297539.89 |================================================ GCC 10.0.0 znver2 . 298969.75 |================================================ GCC 10.0.0 ........ 300244.81 |================================================ PostgreSQL pgbench 10.3 Scaling: Buffer Test - Test: Normal Load - Mode: Read Write TPS > Higher Is Better GCC 9.1.0 ......... 29178.23 |================================================= GCC 9.1.0 znver2 .. 29149.20 |================================================= GCC 10.0.0 znver2 . 29148.60 |================================================= GCC 10.0.0 ........ 29372.39 |================================================= Redis 4.0.8 Test: GET Requests Per Second > Higher Is Better GCC 9.1.0 ......... 3297713.33 |=============================================== GCC 9.1.0 znver2 .. 3066070.28 |============================================ GCC 10.0.0 znver2 . 3031706.22 |=========================================== GCC 10.0.0 ........ 3042507.47 |=========================================== Redis 4.0.8 Test: SET Requests Per Second > Higher Is Better GCC 9.1.0 ......... 2122162.94 |============================================== GCC 9.1.0 znver2 .. 2169531.00 |=============================================== GCC 10.0.0 znver2 . 2084989.88 |============================================= GCC 10.0.0 ........ 2051361.33 |============================================ SciMark 2.0 Computational Test: Composite Mflops > Higher Is Better GCC 9.1.0 ......... 2768.16 |====================================== GCC 9.1.0 znver2 .. 3686.60 |================================================== GCC 10.0.0 znver2 . 3553.67 |================================================ GCC 10.0.0 ........ 3127.49 |========================================== SciMark 2.0 Computational Test: Monte Carlo Mflops > Higher Is Better GCC 9.1.0 ......... 761.38 |================================================= GCC 9.1.0 znver2 .. 800.23 |=================================================== GCC 10.0.0 znver2 . 759.97 |================================================ GCC 10.0.0 ........ 777.17 |================================================== SciMark 2.0 Computational Test: Fast Fourier Transform Mflops > Higher Is Better GCC 9.1.0 ......... 295.18 |================================================== GCC 9.1.0 znver2 .. 273.49 |============================================== GCC 10.0.0 znver2 . 261.10 |============================================ GCC 10.0.0 ........ 301.17 |=================================================== SciMark 2.0 Computational Test: Sparse Matrix Multiply Mflops > Higher Is Better GCC 9.1.0 ......... 3767.63 |================================================= GCC 9.1.0 znver2 .. 3580.73 |============================================== GCC 10.0.0 znver2 . 3675.94 |================================================ GCC 10.0.0 ........ 3856.63 |================================================== SciMark 2.0 Computational Test: Dense LU Matrix Factorization Mflops > Higher Is Better GCC 9.1.0 ......... 6891.37 |============================== GCC 9.1.0 znver2 .. 11370.27 |================================================= GCC 10.0.0 znver2 . 10777.88 |============================================== GCC 10.0.0 ........ 8526.66 |===================================== SciMark 2.0 Computational Test: Jacobi Successive Over-Relaxation Mflops > Higher Is Better GCC 9.1.0 ......... 2125.26 |============================================ GCC 9.1.0 znver2 .. 2408.26 |================================================== GCC 10.0.0 znver2 . 2293.46 |================================================ GCC 10.0.0 ........ 2175.85 |============================================= Smallpt 1.0 Global Illumination Renderer; 128 Samples Seconds < Lower Is Better GCC 9.1.0 ......... 7.78 |===================================================== GCC 9.1.0 znver2 .. 7.67 |==================================================== GCC 10.0.0 znver2 . 7.53 |=================================================== GCC 10.0.0 ........ 7.84 |===================================================== Sockperf 3.4 Test: Throughput Messages Per Second > Higher Is Better GCC 9.1.0 ......... 514551 |================================================== GCC 9.1.0 znver2 .. 517095 |================================================== GCC 10.0.0 znver2 . 529657 |=================================================== GCC 10.0.0 ........ 514748 |================================================== Sockperf 3.4 Test: Latency Ping Pong usec < Lower Is Better GCC 9.1.0 ......... 3.12 |==================================================== GCC 9.1.0 znver2 .. 3.15 |===================================================== GCC 10.0.0 znver2 . 3.04 |=================================================== GCC 10.0.0 ........ 3.03 |=================================================== Stockfish 9 Total Time Nodes Per Second > Higher Is Better GCC 9.1.0 ......... 39278964 |================================================= GCC 9.1.0 znver2 .. 39561655 |================================================= GCC 10.0.0 znver2 . 39540328 |================================================= GCC 10.0.0 ........ 39631993 |================================================= SVT-AV1 0.5 1080p 8-bit YUV To AV1 Video Encode Frames Per Second > Higher Is Better GCC 9.1.0 ......... 46.45 |==================================================== GCC 9.1.0 znver2 .. 46.39 |==================================================== GCC 10.0.0 znver2 . 46.49 |==================================================== GCC 10.0.0 ........ 46.22 |==================================================== SVT-HEVC 2019-02-03 1080p 8-bit YUV To HEVC Video Encode Frames Per Second > Higher Is Better GCC 9.1.0 ......... 246.01 |================================================== GCC 9.1.0 znver2 .. 247.33 |=================================================== GCC 10.0.0 znver2 . 247.99 |=================================================== GCC 10.0.0 ........ 248.85 |=================================================== SVT-VP9 2019-02-17 1080p 8-bit YUV To VP9 Video Encode Frames Per Second > Higher Is Better GCC 9.1.0 ......... 89.99 |================================================ GCC 9.1.0 znver2 .. 96.54 |==================================================== GCC 10.0.0 znver2 . 92.35 |================================================== GCC 10.0.0 ........ 89.84 |================================================ Timed LLVM Compilation 6.0.1 Time To Compile Seconds < Lower Is Better GCC 9.1.0 ......... 280.27 |================================================ GCC 9.1.0 znver2 .. 284.10 |================================================ GCC 10.0.0 znver2 . 300.31 |=================================================== GCC 10.0.0 ........ 292.53 |================================================== Timed PHP Compilation 7.1.9 Time To Compile Seconds < Lower Is Better GCC 9.1.0 ......... 52.71 |================================================== GCC 9.1.0 znver2 .. 53.91 |==================================================== GCC 10.0.0 znver2 . 53.76 |=================================================== GCC 10.0.0 ........ 54.43 |==================================================== TSCP 1.81 AI Chess Performance Nodes Per Second > Higher Is Better GCC 9.1.0 ......... 1305781 |============================================== GCC 9.1.0 znver2 .. 1337188 |=============================================== GCC 10.0.0 znver2 . 1408752 |================================================== GCC 10.0.0 ........ 1366017 |================================================ x264 2018-09-25 H.264 Video Encoding Frames Per Second > Higher Is Better GCC 9.1.0 ......... 139.59 |=================================================== GCC 9.1.0 znver2 .. 138.41 |================================================== GCC 10.0.0 znver2 . 139.82 |=================================================== GCC 10.0.0 ........ 138.74 |=================================================== x265 3.0 H.265 1080p Video Encoding Frames Per Second > Higher Is Better GCC 9.1.0 ......... 52.94 |==================================================== GCC 9.1.0 znver2 .. 52.53 |==================================================== GCC 10.0.0 znver2 . 52.40 |=================================================== GCC 10.0.0 ........ 53.00 |==================================================== XZ Compression 5.2.4 Compressing ubuntu-16.04.3-server-i386.img, Compression Level 9 Seconds < Lower Is Better GCC 9.1.0 ......... 25.23 |==================================================== GCC 9.1.0 znver2 .. 25.25 |==================================================== GCC 10.0.0 znver2 . 25.39 |==================================================== GCC 10.0.0 ........ 25.26 |====================================================