performance:noctua-fayalite-fist
This is an old revision of the document!
Performance of the Fayalite-FIST Benchmark on Noctua
| Nodes | Cores | Best time (s) | Config |
|---|---|---|---|
| 1 | 40 | 169.698 | 1 OMP thread per MPI task |
| 2 | 80 | 150.330 | 1 OMP thread per MPI task |
| 4 | 160 | 141.707 | 1 OMP thread per MPI task |
| 8 | 320 | 139.177 | 1 OMP thread per MPI task |
| 16 | 640 | 133.411 | 10 OMP threads per MPI task |
| 32 | 1280 | 133.411 | 20 OMP threads per MPI task |
| 64 | 2560 | 119.820 | 10 OMP threads per MPI task |
| 128 | 5120 | 124.026 | 5 OMP threads per MPI task |
| 256 | 10240 | 121.857 | 10 OMP threads per MPI task |
performance/noctua-fayalite-fist.1571392443.txt.gz · Last modified: 2020/08/21 10:15 (external edit)