Improve packing kernel launch efficiency for pipelined backends using CUDA graphs.#68

Merged

romerojosh merged 8 commits intomainfrom

Apr 16, 2025

Commits on Apr 10, 2025

Commits on Apr 11, 2025

Commits on Apr 14, 2025

Fix issue with graph cache clearing during autotuning. Need to clear between each pdim/backend combination.
romerojosh
committed

Commits on Apr 16, 2025

Add datatype to graph cache key.
romerojosh
committed