Table 2.
Top five longest running kernels before and after merging
| Before merging | After merging | ||||||
|---|---|---|---|---|---|---|---|
| Name | Calls | Tot time | Avg time | Name | Calls | Tot time | Avg time |
| MatVec | 3808 | 110.69 ms | 29.067 us | MatVec | 3808 | 110.59 ms | 29.040 us |
| elax1 | 2464 | 55.326 ms | 22.453 us | Relax1 | 2464 | 55.350 ms | 22.463 us |
| Relax0 | 2352 | 45.153 ms | 19.197 us | Relax0 | 2352 | 44.987 ms | 19.126 us |
| InitComm | 20656 | 38.544 ms | 1.8650 us | Axpy | 1660 | 35.664 ms | 21.484 us |
| inComm | 20688 | 37.894 ms | 1.8310 us | Memcpy-HtoD | 12862 | 26.689 ms | 2.0750 us |