Optimized PTX IntrinsicMath implementation to use LibDevice. #5333
Job | Run time |
---|---|
12s | |
11s | |
2s | |
4m 31s | |
2m 0s | |
3m 26s | |
5s | |
56s | |
3m 47s | |
4m 6s | |
4m 14s | |
4m 8s | |
3m 58s | |
4m 9s | |
12m 30s | |
13m 5s | |
12m 30s | |
1m 2s | |
1m 4s | |
58s | |
4m 13s | |
5m 32s | |
4m 40s | |
5m 59s | |
5m 0s | |
6m 2s | |
7m 8s | |
7m 40s | |
7m 30s | |
2m 25s | |
2m 54s | |
2m 42s | |
10m 17s | |
10m 54s | |
10m 40s | |
18m 30s | |
17m 15s | |
17m 19s | |
4m 5s | |
1s | |
3m 23s | |
1s | |
0s | |
0s | |
3h 51m 4s |