Design for runtime compilation of kernel templates with dynamically supplied types
Open
- Dominant language
- C++
- Stars
- 2.5k
- Forks
- 486
- Avg merge
- 2d 6h
- Merged PRs (30d)
- 295
Description
```
struct kernel_functor{
template
__device__ void operator()(){ // kernel implementation }
};
...
cudax::jit_and_launch(kernel_functor, "Type1", "Type2")
```
Is there a way to make the above or something similar work with NVRTC?
Contributor guide
Assessment
This issue has not been assessed yet.