8380 sun 2 x Intel Xeon Platinum 8380 testing with a Intel M50CYP2SB2U (SE5C6200.86B.0022.D08.2103221623 BIOS) and ASPEED on Ubuntu 20.04 via the Phoronix Test Suite. A: Processor: 2 x Intel Xeon Platinum 8380 @ 3.40GHz (80 Cores / 160 Threads), Motherboard: Intel M50CYP2SB2U (SE5C6200.86B.0022.D08.2103221623 BIOS), Chipset: Intel Device 0998, Memory: 512GB, Disk: 3841GB Micron_9300_MTFDHAL3T8TDP, Graphics: ASPEED, Monitor: VE228, Network: 2 x Intel X710 for 10GBASE-T + 2 x Intel E810-C for QSFP OS: Ubuntu 20.04, Kernel: 5.15.11-051511-generic (x86_64), Desktop: GNOME Shell 3.36.9, Display Server: X Server 1.20.13, Vulkan: 1.0.2, Compiler: GCC 9.3.0 + Clang 10.0.0-4ubuntu1, File-System: ext4, Screen Resolution: 1920x1080 B: Processor: 2 x Intel Xeon Platinum 8380 @ 3.40GHz (80 Cores / 160 Threads), Motherboard: Intel M50CYP2SB2U (SE5C6200.86B.0022.D08.2103221623 BIOS), Chipset: Intel Device 0998, Memory: 512GB, Disk: 3841GB Micron_9300_MTFDHAL3T8TDP, Graphics: ASPEED, Monitor: VE228, Network: 2 x Intel X710 for 10GBASE-T + 2 x Intel E810-C for QSFP OS: Ubuntu 20.04, Kernel: 5.15.11-051511-generic (x86_64), Desktop: GNOME Shell 3.36.9, Display Server: X Server 1.20.13, Vulkan: 1.0.2, Compiler: GCC 9.3.0 + Clang 10.0.0-4ubuntu1, File-System: ext4, Screen Resolution: 1920x1080 C: Processor: 2 x Intel Xeon Platinum 8380 @ 3.40GHz (80 Cores / 160 Threads), Motherboard: Intel M50CYP2SB2U (SE5C6200.86B.0022.D08.2103221623 BIOS), Chipset: Intel Device 0998, Memory: 512GB, Disk: 3841GB Micron_9300_MTFDHAL3T8TDP, Graphics: ASPEED, Monitor: VE228, Network: 2 x Intel X710 for 10GBASE-T + 2 x Intel E810-C for QSFP OS: Ubuntu 20.04, Kernel: 5.15.11-051511-generic (x86_64), Desktop: GNOME Shell 3.36.9, Display Server: X Server 1.20.13, Vulkan: 1.0.2, Compiler: GCC 9.3.0 + Clang 10.0.0-4ubuntu1, File-System: ext4, Screen Resolution: 1920x1080 D: Processor: 2 x Intel Xeon Platinum 8380 @ 3.40GHz (80 Cores / 160 Threads), Motherboard: Intel M50CYP2SB2U (SE5C6200.86B.0022.D08.2103221623 BIOS), Chipset: Intel Device 0998, Memory: 512GB, Disk: 3841GB Micron_9300_MTFDHAL3T8TDP, Graphics: ASPEED, Monitor: VE228, Network: 2 x Intel X710 for 10GBASE-T + 2 x Intel E810-C for QSFP OS: Ubuntu 20.04, Kernel: 5.15.11-051511-generic (x86_64), Desktop: GNOME Shell 3.36.9, Display Server: X Server 1.20.13, Vulkan: 1.0.2, Compiler: GCC 9.3.0 + Clang 10.0.0-4ubuntu1, File-System: ext4, Screen Resolution: 1920x1080 perf-bench Benchmark: Epoll Wait ops/sec > Higher Is Better A . 3190 |=============================================================== B . 3376 |================================================================== C . 3518 |===================================================================== D . 3345 |================================================================== perf-bench Benchmark: Futex Hash ops/sec > Higher Is Better A . 2990938 |================================================================== B . 2990356 |================================================================== C . 2985843 |================================================================== D . 2984853 |================================================================== perf-bench Benchmark: Memcpy 1MB GB/sec > Higher Is Better A . 16.85 |==================================================================== B . 15.93 |================================================================ C . 16.27 |================================================================== D . 15.95 |================================================================ perf-bench Benchmark: Memset 1MB GB/sec > Higher Is Better A . 58.85 |=================================================================== B . 53.17 |============================================================ C . 58.45 |================================================================== D . 59.79 |==================================================================== perf-bench Benchmark: Sched Pipe ops/sec > Higher Is Better A . 157484 |================================================================== B . 159641 |=================================================================== C . 159064 |=================================================================== D . 158896 |=================================================================== perf-bench Benchmark: Futex Lock-Pi ops/sec > Higher Is Better A . 47 |=================================================================== B . 44 |============================================================== C . 49 |====================================================================== D . 50 |======================================================================= perf-bench Benchmark: Syscall Basic ops/sec > Higher Is Better A . 13955440 |================================================================= B . 13985008 |================================================================= C . 13976056 |================================================================= D . 13925517 |================================================================= oneDNN 2.6 Harness: IP Shapes 1D - Data Type: f32 - Engine: CPU ms < Lower Is Better A . 0.874369 |================================================================= B . 0.858012 |================================================================ C . 0.863419 |================================================================ D . 0.855213 |================================================================ oneDNN 2.6 Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU ms < Lower Is Better A . 1.27751 |================================================================= B . 1.28130 |================================================================== C . 1.29049 |================================================================== D . 1.28979 |================================================================== oneDNN 2.6 Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better A . 1.30484 |================================================================== B . 1.29761 |================================================================== C . 1.28310 |================================================================= D . 1.27539 |================================================================= oneDNN 2.6 Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better A . 0.444806 |================================================================= B . 0.438798 |================================================================ C . 0.443744 |================================================================= D . 0.445376 |================================================================= oneDNN 2.6 Harness: IP Shapes 1D - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better A . 2.98315 |================================================================== B . 3.00061 |================================================================== C . 2.99363 |================================================================== D . 2.99408 |================================================================== oneDNN 2.6 Harness: IP Shapes 3D - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better A . 1.81335 |================================================================== B . 1.81013 |================================================================== C . 1.82023 |================================================================== D . 1.82082 |================================================================== oneDNN 2.6 Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU ms < Lower Is Better A . 1.38472 |================================================================== B . 1.39272 |================================================================== C . 1.38829 |================================================================== D . 1.38183 |================================================================= oneDNN 2.6 Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPU ms < Lower Is Better A . 7.24738 |================================================================== B . 7.23701 |================================================================= C . 7.26969 |================================================================== D . 7.29653 |================================================================== oneDNN 2.6 Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPU ms < Lower Is Better A . 0.884168 |================================================================= B . 0.888675 |================================================================= C . 0.888288 |================================================================= D . 0.884572 |================================================================= oneDNN 2.6 Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better A . 1.13824 |================================================================= B . 1.12915 |================================================================= C . 1.14503 |================================================================== D . 1.15270 |================================================================== oneDNN 2.6 Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better A . 0.372274 |================================================================= B . 0.369100 |================================================================ C . 0.364478 |================================================================ D . 0.372693 |================================================================= oneDNN 2.6 Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better A . 0.194293 |================================================================ B . 0.196251 |================================================================= C . 0.194502 |================================================================ D . 0.191428 |=============================================================== oneDNN 2.6 Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU ms < Lower Is Better A . 627.39 |=================================================================== B . 625.62 |=================================================================== C . 621.73 |================================================================== D . 622.50 |================================================================== oneDNN 2.6 Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU ms < Lower Is Better A . 379.17 |=================================================================== B . 379.59 |=================================================================== C . 379.63 |=================================================================== D . 378.23 |=================================================================== oneDNN 2.6 Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better A . 618.68 |================================================================== B . 626.98 |=================================================================== C . 618.14 |================================================================== D . 617.79 |================================================================== oneDNN 2.6 Harness: Convolution Batch Shapes Auto - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better A . 2.12309 |================================================================== B . 2.11786 |================================================================== C . 2.10492 |================================================================= D . 2.11735 |================================================================== oneDNN 2.6 Harness: Deconvolution Batch shapes_1d - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better A . 3.78575 |================================================================== B . 3.79039 |================================================================== C . 3.76245 |================================================================== D . 3.77449 |================================================================== oneDNN 2.6 Harness: Deconvolution Batch shapes_3d - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better A . 3.62162 |================================================================== B . 3.61967 |================================================================== C . 3.60012 |================================================================== D . 3.60258 |================================================================== oneDNN 2.6 Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better A . 378.86 |================================================================== B . 383.21 |=================================================================== C . 379.21 |================================================================== D . 378.64 |================================================================== oneDNN 2.6 Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU ms < Lower Is Better A . 0.230657 |================================================================ B . 0.234187 |================================================================= C . 0.232623 |================================================================= D . 0.232562 |================================================================= oneDNN 2.6 Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better A . 636.29 |=================================================================== B . 621.88 |================================================================= C . 618.82 |================================================================= D . 617.73 |================================================================= oneDNN 2.6 Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better A . 378.96 |================================================================== B . 386.18 |=================================================================== C . 379.59 |================================================================== D . 377.23 |================================================================= oneDNN 2.6 Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better A . 0.177939 |================================================================= B . 0.170109 |============================================================== C . 0.172350 |=============================================================== D . 0.172182 |=============================================================== oneDNN 2.6 Harness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better A . 2.08241 |================================================================== B . 2.09074 |================================================================== C . 2.05510 |================================================================= D . 2.08466 |================================================================== Java JMH Throughput Ops/s > Higher Is Better A . 117974431957.74 |========================================================== B . 117397403601.04 |========================================================== C . 117539633814.56 |========================================================== D . 117793385377.77 |==========================================================