I was sad to learn that one cannot use dynamic parallelism in cooperative kernel launches. Is it a hardware limitation or a software one that may be lifted in the future?
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Can I use cooperative_groups::sync(grid) in child kernel (CUDA dynamic parallelism)? | 1 | 492 | January 24, 2022 | |
| Dynamic Parallelism | 2 | 376 | February 7, 2018 | |
| Cuda-task parallelism on a single GPU | 0 | 504 | October 15, 2020 | |
| Dynamic Parallelism extreme slowdown | 0 | 836 | April 7, 2013 | |
| Dynamic Parallelism and cudaGetLastError() | 1 | 779 | April 4, 2013 | |
| Parallel execution of kernels C# | 1 | 824 | July 26, 2011 | |
| dynamic parallelism performance | 4 | 954 | January 3, 2013 | |
| Dynamic Parallelism in OptiX | 1 | 571 | April 20, 2023 | |
| Dynamic parallelism vs flat kernels | 0 | 369 | May 30, 2017 | |
| task parallelism on CUDA | 3 | 1737 | June 24, 2012 |