[FEA]: Tune `DeviceReduce::Sum/Min` for RTX A6000, H100, B200 to test new warp-reduce optimizations
Open
- Dominant language
- C++
- Stars
- 2.5k
- Forks
- 486
- Avg merge
- 2d 6h
- Merged PRs (30d)
- 295
Description
Tune `DeviceReduce::Sum/Min` to observe performance changes for https://github.com/NVIDIA/cccl/pull/4312
Contributor guide
Assessment
This issue has not been assessed yet.