AMD EPYC Compiler Testing July 2018 AMD EPYC compiler benchmarks. GCC 8.2 RC1: -O3 -march=native: Processor: AMD EPYC 7601 32-Core @ 2.20GHz (32 Cores / 64 Threads), Motherboard: TYAN B8026T70AE24HR (V1.02.B10 BIOS), Chipset: AMD Family 17h, Memory: 129024MB, Disk: 280GB INTEL SSDPE21D280GA, Graphics: ASPEED ASPEED Family, Monitor: VE228, Network: Broadcom Limited NetXtreme BCM5720 Gigabit PCIe OS: Ubuntu 18.10, Kernel: 4.16.0-041600-generic (x86_64), Compiler: GCC 8.1.1 20180720, File-System: ext4, Screen Resolution: 1920x1080 AMD EPYC 7601 32-Core: Processor: AMD EPYC 7601 32-Core @ 2.20GHz (32 Cores / 64 Threads), Motherboard: TYAN B8026T70AE24HR (V1.02.B10 BIOS), Chipset: AMD Family 17h, Memory: 129024MB, Disk: 280GB INTEL SSDPE21D280GA, Graphics: ASPEED ASPEED Family, Monitor: VE228, Network: Broadcom Limited NetXtreme BCM5720 Gigabit PCIe OS: Ubuntu 18.10, Kernel: 4.16.0-041600-generic (x86_64), Compiler: GCC 8.1.1 20180720, File-System: ext4, Screen Resolution: 1920x1080 SciMark 2.0 Computational Test: Composite Mflops > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 1943.30 |===================================== AMD EPYC 7601 32-Core .......... 1766.38 |================================== SciMark 2.0 Computational Test: Fast Fourier Transform Mflops > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 219.38 |====================================== AMD EPYC 7601 32-Core .......... 214.57 |===================================== SciMark 2.0 Computational Test: Jacobi Successive Over-Relaxation Mflops > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 1687.64 |===================================== AMD EPYC 7601 32-Core .......... 1676.49 |===================================== SciMark 2.0 Computational Test: Monte Carlo Mflops > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 556.37 |====================================== AMD EPYC 7601 32-Core .......... 195.98 |============= SciMark 2.0 Computational Test: Sparse Matrix Multiply Mflops > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 2421.38 |===================================== AMD EPYC 7601 32-Core .......... 2442.87 |===================================== SciMark 2.0 Computational Test: Dense LU Matrix Factorization Mflops > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 4831.75 |===================================== AMD EPYC 7601 32-Core .......... 4301.99 |================================= VP9 libvpx Encoding 1.7.0 vpxenc Frames Per Second > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 12.38 |======================================= FFTW 3.3.6 Build: Stock - Size: 1D FFT Size 1024 Mflops > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 8518.67 |===================================== AMD EPYC 7601 32-Core .......... 8289.10 |==================================== FFTW 3.3.6 Build: Stock - Size: 2D FFT Size 1024 Mflops > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 6557.83 |===================================== AMD EPYC 7601 32-Core .......... 6351.50 |==================================== PolyBench-C 4.2 Test: 3 Matrix Multiplications Seconds < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 3.58 |======================================= AMD EPYC 7601 32-Core .......... 3.66 |======================================== PolyBench-C 4.2 Test: Correlation Computation Seconds < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 5.34 |======================================== AMD EPYC 7601 32-Core .......... 5.39 |======================================== PolyBench-C 4.2 Test: Covariance Computation Seconds < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 5.34 |======================================== AMD EPYC 7601 32-Core .......... 5.37 |======================================== SQLite 3.22 Timed SQLite Insertions Seconds < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 2.89 |===================================== AMD EPYC 7601 32-Core .......... 3.13 |======================================== OpenSSL 1.1.0f RSA 4096-bit Performance Signs Per Second > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 4610.63 |===================================== AMD EPYC 7601 32-Core .......... 4656.40 |===================================== AOBench Size: 2048 x 2048 - Total Time Seconds < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 50.80 |================================== AMD EPYC 7601 32-Core .......... 57.85 |======================================= 7-Zip Compression 16.02 Compress Speed Test MIPS > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 100684 |====================================== AMD EPYC 7601 32-Core .......... 98726 |===================================== Zstd Compression 1.3.4 Compressing ubuntu-16.04.3-server-i386.img, Compression Level 19 Seconds < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 155.64 |======================= AMD EPYC 7601 32-Core .......... 260.02 |====================================== m-queens 1.1 Time To Solve Seconds < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 32.02 | AMD EPYC 7601 32-Core .......... 2694.47 |===================================== C-Ray 1.1 Total Time Seconds < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 2.52 | AMD EPYC 7601 32-Core .......... 566.22 |====================================== Crafty 25.2 Elapsed Time Nodes Per Second > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 6014866 |===================================== AMD EPYC 7601 32-Core .......... 5998759 |===================================== Smallpt 1.0 Global Illumination Renderer; 100 Samples Seconds < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 4 | AMD EPYC 7601 32-Core .......... 533 |========================================= Stockfish 9 Total Time Nodes Per Second > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 59730290 |==================================== AMD EPYC 7601 32-Core .......... 1372510 |= TSCP 1.81 AI Chess Performance Nodes Per Second > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 851019 |====================================== AMD EPYC 7601 32-Core .......... 729507 |================================= Redis 4.0.8 Test: SET Requests Per Second > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 1375161.96 |================================== Redis 4.0.8 Test: GET Requests Per Second > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 1691908.00 |================================== Redis 4.0.8 Test: LPUSH Requests Per Second > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 1108304.54 |================================== Redis 4.0.8 Test: LPOP Requests Per Second > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 1108268.08 |================================== Redis 4.0.8 Test: SADD Requests Per Second > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 1430638.45 |================================== PostgreSQL pgbench 10.3 Scaling: Buffer Test - Test: Single Thread - Mode: Read Write TPS > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 2071.78 |===================================== AMD EPYC 7601 32-Core .......... 1427.38 |========================= PostgreSQL pgbench 10.3 Scaling: Buffer Test - Test: Single Thread - Mode: Read Only TPS > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 14374.13 |==================================== AMD EPYC 7601 32-Core .......... 9213.45 |======================= PostgreSQL pgbench 10.3 Scaling: Buffer Test - Test: Normal Load - Mode: Read Write TPS > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 36226.69 |==================================== AMD EPYC 7601 32-Core .......... 28411.49 |============================ PostgreSQL pgbench 10.3 Scaling: Buffer Test - Test: Normal Load - Mode: Read Only TPS > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 408873.25 |=================================== AMD EPYC 7601 32-Core .......... 242221.35 |===================== libjpeg-turbo tjbench 1.5.3 Test: Decompression Throughput Megapixels/sec > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 142.18 |====================================== AMD EPYC 7601 32-Core .......... 106.03 |============================ Apache Benchmark 2.4.29 Static Web Page Serving Requests Per Second > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 16205.32 |==================================== AMD EPYC 7601 32-Core .......... 15210.47 |================================== LAME MP3 Encoding 3.100 WAV To MP3 Seconds < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 10.51 |=========================== AMD EPYC 7601 32-Core .......... 15.40 |======================================= FLAC Audio Encoding 1.3.2 WAV To FLAC Seconds < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 12.63 |====================================== AMD EPYC 7601 32-Core .......... 12.87 |======================================= ebizzy 0.3 Records/s > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 1147009 |===================================== AMD EPYC 7601 32-Core .......... 1131696 |===================================== GraphicsMagick 1.3.28 Operation: HWB Color Space Iterations Per Minute > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 209 |========================================= AMD EPYC 7601 32-Core .......... 46 |========= GraphicsMagick 1.3.28 Operation: Blur Iterations Per Minute > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 154 |========================================= AMD EPYC 7601 32-Core .......... 18 |===== GraphicsMagick 1.3.28 Operation: Local Adaptive Thresholding Iterations Per Minute > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 110 |========================================= AMD EPYC 7601 32-Core .......... 23 |========= GraphicsMagick 1.3.28 Operation: Resizing Iterations Per Minute > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 185 |========================================= AMD EPYC 7601 32-Core .......... 29 |====== GraphicsMagick 1.3.28 Operation: Sharpen Iterations Per Minute > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 188 |========================================= AMD EPYC 7601 32-Core .......... 6 |= Timed HMMer Search 2.3.2 Pfam Database Search Seconds < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 7.65 |== AMD EPYC 7601 32-Core .......... 192.26 |====================================== Bullet Physics Engine 2.81 Test: 3000 Fall Seconds < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 5.02 |======================================== Bullet Physics Engine 2.81 Test: 1000 Stack Seconds < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 5.79 |======================================== Bullet Physics Engine 2.81 Test: 136 Ragdolls Seconds < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 3.09 |======================================== Bullet Physics Engine 2.81 Test: 1000 Convex Seconds < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 5.32 |======================================== Bullet Physics Engine 2.81 Test: Prim Trimesh Seconds < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 1.13 |======================================== Bullet Physics Engine 2.81 Test: Convex Trimesh Seconds < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 1.33 |======================================== Bullet Physics Engine 2.81 Test: Raytests Seconds < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 3.11 |======================================== TTSIOD 3D Renderer 2.3a Phong Rendering With Soft-Shadow Mapping FPS > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 439.54 |====================================== AMD EPYC 7601 32-Core .......... 3.04 | Himeno Benchmark 3.0 Poisson Pressure Solver MFLOPS > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 982.67 |====================================== AMD EPYC 7601 32-Core .......... 972.07 |====================================== Tinymembench 2018-05-28 Standard Memcpy MB/s > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 12497.20 |==================================== AMD EPYC 7601 32-Core .......... 12463.50 |==================================== Tinymembench 2018-05-28 Standard Memset MB/s > Higher Is Better GCC 8.2 RC1: -O3 -march=native . 10063.10 |==================================== AMD EPYC 7601 32-Core .......... 10015.50 |==================================== Sockperf 3.4 Test: Latency Under Load usec < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 19.75 |========================= AMD EPYC 7601 32-Core .......... 31.29 |======================================= Sockperf 3.4 Test: Latency Ping Pong usec < Lower Is Better GCC 8.2 RC1: -O3 -march=native . 6.42 |===================================== AMD EPYC 7601 32-Core .......... 6.88 |========================================