Basically, my program is fine and it runs properly.
The problem arises when I want to speed it up by setting “Generate GPU Debug Information” to “No”, indeed I get error until I am not below 512 kernels…
Why?
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Kernel crash when GPU Debug Info is disabled in Visual Studio | 5 | 1106 | March 12, 2018 | |
| Get error cudaErrorLaunchOutOfResources from kernel after removing -G compiler option | 1 | 1282 | March 10, 2019 | |
| Cannot debug inside GPU | 4 | 1217 | September 8, 2010 | |
| cuda-gdb and cudaCheckError() too many resources requested for launch | 1 | 598 | March 22, 2019 | |
| Cuda-gdb doesn't break and/or step into Kernels | 26 | 54529 | August 1, 2011 | |
| Limited output of kernel printf | 4 | 2464 | September 1, 2011 | |
| How to debug kernel throwing an exception? | 16 | 8291 | June 14, 2013 | |
| Building with "-g -G" reduces max blocks per SM | 0 | 556 | October 5, 2022 | |
| cuda-gdb won't print local variables in kernel function | 12 | 18052 | July 7, 2011 | |
| Debugger doesn't break in device code | 3 | 971 | October 9, 2013 |