you may be hitting a lazy loading situation
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| How to overlap execution of kernels in different streams with copy operations | 9 | 1007 | February 1, 2022 | |
| Concurrent copy & execution problem Device to host memory copy is not overlapped with kernel exe | 1 | 1781 | June 23, 2010 | |
| About Stream control | 1 | 949 | March 26, 2009 | |
| cuda stream | 3 | 5819 | April 6, 2011 | |
| Syncronization with cuda Streams | 8 | 435 | October 12, 2021 | |
| Ordering of cudaMemcpyAsync issued to separate streams | 4 | 599 | February 5, 2019 | |
| cudaMemcpyAsync | 10 | 21247 | October 16, 2015 | |
| Overlap cudaMemcpyAsync and kernel | 1 | 514 | February 10, 2021 | |
| Asynchronous cudaMemcpy host to device not overlapping with kernel | 4 | 54 | April 10, 2025 | |
| Memory copy/set async to kernel execution in different stream | 5 | 1126 | December 15, 2022 |