Table 3 Comparison of runtimes (in seconds) for processing the nne 4096_911 dataset with various parameters.
From: Parallel mutual information estimation for inferring gene regulatory networks on GPUs
Bin Number | Spline Order | MIBE (4 threads) | CUDA-MI | Speedup |
|---|---|---|---|---|
10 | 3 | 867.35 | 45.17 | 19.2 |
10 | 4 | 867.24 | 46.18 | 18.8 |
10 | 5 | 868.3 | 46.54 | 18.7 |
15 | 3 | 1926.67 | 60.02 | 32.1 |
15 | 4 | 1947.45 | 60.83 | 32.0 |
15 | 5 | 1949.65 | 61.77 | 31.6 |
20 | 3 | 3468.07 | 104.45 | 33.2 |
20 | 4 | 3657.64 | 105.40 | 34.7 |
20 | 5 | 3658.83 | 105.61 | 34.6 |