Optimized PTX IntrinsicMath implementation to use LibDevice. #5039
Job | Run time |
---|---|
9s | |
2m 25s | |
5m 27s | |
4m 28s | |
1m 27s | |
10s | |
4s | |
4s | |
3m 46s | |
4m 9s | |
4m 36s | |
5m 18s | |
4m 47s | |
4m 8s | |
6m 55s | |
9m 6s | |
7m 47s | |
1m 20s | |
1m 2s | |
1m 0s | |
5m 38s | |
6m 47s | |
7m 15s | |
6m 18s | |
6m 30s | |
10m 2s | |
6m 52s | |
8m 6s | |
7m 47s | |
4m 31s | |
4m 54s | |
3m 42s | |
8m 56s | |
9m 14s | |
9m 5s | |
39m 52s | |
40m 44s | |
41m 42s | |
6m 6s | |
1s | |
5m 57s | |
1s | |
0s | |
0s | |
5h 8m 8s |