Table 2 Inference processing time on different hardware (ms/example).
From: Deep Cytometry: Deep learning with Real-time Inference in Cell Sorting and Flow Cytometry
Batch size (no. of examples) | Intel Xeon CPU | NVIDIA K80 GPU | NVIDIA P100 GPU |
|---|---|---|---|
1 | 23.2 | 8.6 | 3.6 |
2 | 21.0 | 5.7 | 2.0 |
4 | 12.9 | 4.3 | 1.6 |
8 | 10.0 | 3.7 | 1.3 |
16 | 8.7 | 2.9 | 0.8 |
32 | 8.0 | 2.6 | 0.7 |
64 | 7.4 | 1.9 | 0.6 |
