[EXP][CMDBUF] Improve CUDA Fill op implementation #5381
Job | Run time |
---|---|
0s | |
3m 47s | |
4m 9s | |
6m 57s | |
3m 46s | |
4m 23s | |
3m 21s | |
4m 46s | |
2m 56s | |
4m 39s | |
8m 52s | |
7m 43s | |
10m 45s | |
9m 22s | |
10m 11s | |
11m 18s | |
8m 53s | |
8m 14s | |
11m 33s | |
9m 33s | |
10m 12s | |
11m 33s | |
7m 19s | |
8m 46s | |
14m 9s | |
9m 15s | |
11m 41s | |
4m 2s | |
7m 29s | |
13m 44s | |
4m 18s | |
9m 10s | |
4m 32s | |
4m 23s | |
7m 6s | |
4m 51s | |
6m 7s | |
3m 2s | |
3m 57s | |
4m 1s | |
8m 7s | |
5m 52s | |
2m 53s | |
2m 28s | |
5m 59s | |
5h 10m 4s |