cudev: Add __shfl_down implementation for long long and unsigned long for CUDA Tookit < 9.0 #1391
Job | Run time |
---|---|
0s | |
2s | |
1h 5m 58s | |
1h 2m 31s | |
14m 26s | |
1h 33m 7s | |
56m 17s | |
39s | |
49m 38s | |
28m 56s | |
30m 1s | |
30m 39s | |
7h 12m 14s |
Job | Run time |
---|---|
0s | |
2s | |
1h 5m 58s | |
1h 2m 31s | |
14m 26s | |
1h 33m 7s | |
56m 17s | |
39s | |
49m 38s | |
28m 56s | |
30m 1s | |
30m 39s | |
7h 12m 14s |