Table 10 Comparison of Learning and Inference Speed on GPUs
From: DASSI: differential architecture search for splice identification from DNA sequences
GPU Model | Training | Inference | ||||
|---|---|---|---|---|---|---|
| Â | Single | Half | Double | Single | Half | Double |
Quadro K4100M | 1041.24 | 1090.38 | 1902.38 | 134.97 | 124.59 | 274.78 |
Tesla K40m | 352.53 | 470.72 | 485.17 | 48.022 | 49.38 | 62.13 |
Tesla V100-SXM2 | 68.46 | 119.55 | 130.09 | 13.21 | 11.02 | 11.07 |