Skip to content

[Common] Reduce shared-memory bank conflicts in the colwise scaling path of the tuned NVFP4 kernel#3106

Merged
ptrendx merged 1 commit into
NVIDIA:mainfrom
Oleg-Goncharov:pr_nvfp4_micro_optimization
Jun 9, 2026
Merged

[Common] Reduce shared-memory bank conflicts in the colwise scaling path of the tuned NVFP4 kernel#3106
ptrendx merged 1 commit into
NVIDIA:mainfrom
Oleg-Goncharov:pr_nvfp4_micro_optimization

Commits

Commits on Jun 8, 2026