[FEA] Use `bit_cast` instead of memcpy in cuda::atomics
Open
- Dominant language
- C++
- Stars
- 2.5k
- Forks
- 486
- Avg merge
- 2d 6h
- Merged PRs (30d)
- 295
Description
> would be better to use `cuda::std::bit_cast`
_Originally posted by @fbusato in [#3440](https://github.com/NVIDIA/cccl/pull/3440/files#r2414856722)_
Contributor guide
Assessment
This issue has not been assessed yet.