Are there any strategies for mitigating the slowdown due to dynamic parallelism overhead? I’m currently seeing slow downs of 20x to over 100x just launching a single empty do nothing thread from my host launched kernels.
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Dynamic Parallelism extreme slowdown | 1 | 1059 | April 7, 2013 | |
| a question about low performance on dynamic parallelism with tremendous data | 2 | 1231 | May 27, 2013 | |
| Cuda Dynamic Parallelism Performance | 3 | 2017 | July 14, 2016 | |
| dynamic parallelism performance | 4 | 1036 | January 3, 2013 | |
| Is this strategy not suitable for dynamic parallelism ? | 0 | 520 | January 9, 2014 | |
| How much benefit can i get from dynamic parallelism in my code | 0 | 688 | December 24, 2013 | |
| dynamic parallelism | 3 | 1161 | December 30, 2012 | |
| Performance drops with dynamic parallelism | 12 | 955 | June 3, 2024 | |
| Performance slowdown in Titan V in the presence of Dynamic Parallelism code. | 1 | 516 | October 15, 2018 | |
| GTX Titan and dynamic parallelism | 6 | 1384 | November 14, 2013 |