eMAG Ampere eMAG ARMv8 testing with a AmpereComputing OSPREY (4.8.19 BIOS) and ASPEED on Ubuntu 20.04 via the Phoronix Test Suite. 1: Processor: Ampere eMAG ARMv8 @ 3.00GHz (32 Cores), Motherboard: AmpereComputing OSPREY (4.8.19 BIOS), Chipset: Applied Micro Circuits X-Gene, Memory: 126GB, Disk: 256GB Samsung SSD 860, Graphics: ASPEED, Monitor: VE228, Network: Intel I210 OS: Ubuntu 20.04, Kernel: 5.7.0-050700-generic (aarch64), Desktop: GNOME Shell 3.36.3, Display Server: X Server 1.20.8, Display Driver: modesetting 1.20.8, Compiler: GCC 9.3.0, File-System: ext4, Screen Resolution: 1920x1080 2: Processor: Ampere eMAG ARMv8 @ 3.00GHz (32 Cores), Motherboard: AmpereComputing OSPREY (4.8.19 BIOS), Chipset: Applied Micro Circuits X-Gene, Memory: 126GB, Disk: 256GB Samsung SSD 860, Graphics: ASPEED, Monitor: VE228, Network: Intel I210 OS: Ubuntu 20.04, Kernel: 5.7.0-050700-generic (aarch64), Desktop: GNOME Shell 3.36.3, Display Server: X Server 1.20.8, Display Driver: modesetting 1.20.8, Compiler: GCC 9.3.0, File-System: ext4, Screen Resolution: 1920x1080 3: Processor: Ampere eMAG ARMv8 @ 3.00GHz (32 Cores), Motherboard: AmpereComputing OSPREY (4.8.19 BIOS), Chipset: Applied Micro Circuits X-Gene, Memory: 126GB, Disk: 256GB Samsung SSD 860, Graphics: ASPEED, Monitor: VE228, Network: Intel I210 OS: Ubuntu 20.04, Kernel: 5.7.0-050700-generic (aarch64), Desktop: GNOME Shell 3.36.3, Display Server: X Server 1.20.8, Display Driver: modesetting 1.20.8, Compiler: GCC 9.3.0, File-System: ext4, Screen Resolution: 1920x1080 4: Processor: Ampere eMAG ARMv8 @ 3.00GHz (32 Cores), Motherboard: AmpereComputing OSPREY (4.8.19 BIOS), Chipset: Applied Micro Circuits X-Gene, Memory: 126GB, Disk: 256GB Samsung SSD 860, Graphics: ASPEED, Monitor: VE228, Network: Intel I210 OS: Ubuntu 20.04, Kernel: 5.7.0-050700-generic (aarch64), Desktop: GNOME Shell 3.36.3, Display Server: X Server 1.20.8, Display Driver: modesetting 1.20.8, Compiler: GCC 9.3.0, File-System: ext4, Screen Resolution: 1920x1080 CLOMP 1.2 Static OMP Speedup Speedup > Higher Is Better 1 . 7.1 |===================================================================== 2 . 7.2 |====================================================================== 3 . 7.0 |==================================================================== 4 . 7.2 |====================================================================== Timed MAFFT Alignment 7.471 Multiple Sequence Alignment - LSU RNA Seconds < Lower Is Better 1 . 35.64 |=================================================================== 2 . 35.34 |================================================================== 3 . 36.36 |==================================================================== 4 . 36.00 |=================================================================== simdjson 0.7.1 Throughput Test: Kostya GB/s > Higher Is Better 1 . 0.48 |===================================================================== 2 . 0.48 |===================================================================== 3 . 0.48 |===================================================================== 4 . 0.48 |===================================================================== simdjson 0.7.1 Throughput Test: LargeRandom GB/s > Higher Is Better 1 . 0.23 |===================================================================== 2 . 0.23 |===================================================================== 3 . 0.23 |===================================================================== 4 . 0.23 |===================================================================== simdjson 0.7.1 Throughput Test: PartialTweets GB/s > Higher Is Better 1 . 0.55 |===================================================================== 2 . 0.55 |===================================================================== 3 . 0.55 |===================================================================== 4 . 0.55 |===================================================================== simdjson 0.7.1 Throughput Test: DistinctUserID GB/s > Higher Is Better 1 . 0.56 |==================================================================== 2 . 0.56 |==================================================================== 3 . 0.56 |==================================================================== 4 . 0.57 |===================================================================== TSCP 1.81 AI Chess Performance Nodes Per Second > Higher Is Better 1 . 515903 |=================================================================== 2 . 515903 |=================================================================== 3 . 515710 |=================================================================== 4 . 515709 |=================================================================== oneDNN 2.0 Harness: IP Shapes 1D - Data Type: f32 - Engine: CPU ms < Lower Is Better 1 . 29.01 |=========================================================== 2 . 33.22 |==================================================================== 3 . 32.28 |================================================================== 4 . 30.89 |=============================================================== oneDNN 2.0 Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU ms < Lower Is Better 1 . 23.68 |================================================================== 2 . 22.80 |=============================================================== 3 . 22.85 |================================================================ 4 . 24.45 |==================================================================== oneDNN 2.0 Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better 1 . 183.92 |=================================================================== 2 . 184.88 |=================================================================== 3 . 183.39 |================================================================== 4 . 183.14 |================================================================== oneDNN 2.0 Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better 1 . 418.33 |================================================================== 2 . 421.22 |================================================================== 3 . 424.64 |=================================================================== 4 . 419.39 |================================================================== oneDNN 2.0 Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU ms < Lower Is Better 1 . 93.47 |================================================================= 2 . 98.28 |==================================================================== 3 . 76.35 |===================================================== 4 . 84.37 |========================================================== oneDNN 2.0 Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPU ms < Lower Is Better 1 . 191.84 |================================================================== 2 . 194.62 |=================================================================== 3 . 173.54 |============================================================ 4 . 173.05 |============================================================ oneDNN 2.0 Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPU ms < Lower Is Better 1 . 60.37 |================================================================= 2 . 63.42 |==================================================================== 3 . 55.94 |============================================================ 4 . 59.78 |================================================================ oneDNN 2.0 Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better 1 . 113.02 |=================================================================== 2 . 112.31 |=================================================================== 3 . 112.55 |=================================================================== 4 . 112.92 |=================================================================== oneDNN 2.0 Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better 1 . 184.80 |=================================================================== 2 . 185.97 |=================================================================== 3 . 185.16 |=================================================================== 4 . 183.15 |================================================================== oneDNN 2.0 Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better 1 . 114.07 |=================================================================== 2 . 113.51 |=================================================================== 3 . 113.11 |================================================================== 4 . 112.61 |================================================================== oneDNN 2.0 Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU ms < Lower Is Better 1 . 31838.2 |================================================================== 2 . 30600.5 |=============================================================== 3 . 31978.2 |================================================================== 4 . 30770.0 |================================================================ oneDNN 2.0 Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU ms < Lower Is Better 1 . 17700.4 |================================================================== 2 . 17313.5 |================================================================= 3 . 16864.3 |=============================================================== oneDNN 2.0 Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better 1 . 30113.8 |================================================================ 2 . 30971.0 |================================================================== 3 . 30446.5 |================================================================= oneDNN 2.0 Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better 1 . 16255.9 |================================================================ 2 . 16777.0 |================================================================== 3 . 16436.8 |================================================================= oneDNN 2.0 Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU ms < Lower Is Better 1 . 19.82 |================================================================ 2 . 21.01 |==================================================================== 3 . 20.48 |================================================================== oneDNN 2.0 Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better 1 . 29827.8 |============================================================== 2 . 31520.3 |================================================================== 3 . 30673.7 |================================================================ oneDNN 2.0 Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better 1 . 17140.1 |================================================================== 2 . 16556.7 |================================================================ 3 . 17065.5 |================================================================== oneDNN 2.0 Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better 1 . 36.61 |================================================================= 2 . 38.37 |==================================================================== 3 . 38.49 |==================================================================== rav1e 0.4 Alpha Speed: 1 Frames Per Second > Higher Is Better 1 . 0.084 |==================================================================== 2 . 0.084 |==================================================================== 3 . 0.084 |==================================================================== rav1e 0.4 Alpha Speed: 5 Frames Per Second > Higher Is Better 1 . 0.188 |==================================================================== 2 . 0.187 |==================================================================== 3 . 0.187 |==================================================================== rav1e 0.4 Alpha Speed: 6 Frames Per Second > Higher Is Better 1 . 0.223 |==================================================================== 2 . 0.221 |=================================================================== 3 . 0.222 |==================================================================== rav1e 0.4 Alpha Speed: 10 Frames Per Second > Higher Is Better 1 . 0.425 |==================================================================== 2 . 0.419 |=================================================================== 3 . 0.420 |=================================================================== x264 2019-12-17 H.264 Video Encoding Frames Per Second > Higher Is Better 1 . 32.13 |=================================================================== 2 . 32.67 |==================================================================== 3 . 32.78 |==================================================================== Coremark 1.0 CoreMark Size 666 - Iterations Per Second Iterations/Sec > Higher Is Better 1 . 385397.37 |================================================================ 2 . 385035.21 |================================================================ 3 . 385080.51 |================================================================ Stockfish 12 Total Time Nodes Per Second > Higher Is Better 1 . 15691469 |================================================================= 2 . 15417127 |================================================================ 3 . 15617607 |================================================================= asmFish 2018-07-23 1024 Hash Memory, 26 Depth Nodes/second > Higher Is Better 1 . 33037962 |================================================================= 2 . 33135767 |================================================================= libavif avifenc 0.7.3 Encoder Speed: 0 Seconds < Lower Is Better 1 . 404.74 |=================================================================== 2 . 403.85 |=================================================================== libavif avifenc 0.7.3 Encoder Speed: 2 Seconds < Lower Is Better 1 . 250.17 |=================================================================== 2 . 250.71 |=================================================================== libavif avifenc 0.7.3 Encoder Speed: 8 Seconds < Lower Is Better 1 . 22.30 |==================================================================== 2 . 22.33 |==================================================================== libavif avifenc 0.7.3 Encoder Speed: 10 Seconds < Lower Is Better 1 . 21.88 |==================================================================== 2 . 21.83 |==================================================================== Numpy Benchmark Score > Higher Is Better 1 . 91.65 |==================================================================== 2 . 91.86 |==================================================================== Timed Eigen Compilation 3.3.9 Time To Compile Seconds < Lower Is Better 1 . 357.47 |=================================================================== 2 . 357.14 |=================================================================== Monkey Audio Encoding 3.99.6 WAV To APE Seconds < Lower Is Better 1 . 96.13 |==================================================================== 2 . 39.92 |============================ Opus Codec Encoding 1.3.1 WAV To Opus Encode Seconds < Lower Is Better 1 . 48.31 |==================================================================== 2 . 48.27 |==================================================================== eSpeak-NG Speech Engine 20200907 Text-To-Speech Synthesis Seconds < Lower Is Better 1 . 87.83 |==================================================================== 2 . 84.90 |==================================================================