# gh-6921: Use vectorized transform kernel for sizeof(T) < 4 workloads of arity >1 on Hopper

- Repository: `NVIDIA/cccl`
- URL: https://github.com/NVIDIA/cccl/pull/6921
- Source page: `sources/prs/cccl-cub/PR-6921.md`
- Diff: `review.diff`
- Source snapshot: `1` captured under `source-snapshot/`

See `ORIGIN.yaml` for upstream refs and fetch status.