Optimized PTX IntrinsicMath implementation to use LibDevice. #5369
Job | Run time |
---|---|
10s | |
1s | |
7s | |
1m 36s | |
3m 4s | |
6s | |
2m 9s | |
1s | |
1s | |
1s | |
1s | |
1s | |
1s | |
1s | |
1s | |
1s | |
1s | |
1s | |
1s | |
1s | |
1s | |
1s | |
1s | |
51s | |
53s | |
1m 2s | |
53s | |
50s | |
1m 4s | |
1m 1s | |
55s | |
1s | |
1s | |
1s | |
0s | |
1s | |
1s | |
1s | |
1s | |
15m 5s |