Skip to content

[Pallas] Attention perf: further reduce spillage from pre-loading Q, by loading Q in-loop and not pipelining it#2397

Draft
AmesingFlank wants to merge 1 commit into
mainfrom
AmesingFlank/stack/51
Draft

[Pallas] Attention perf: further reduce spillage from pre-loading Q, by loading Q in-loop and not pipelining it#2397
AmesingFlank wants to merge 1 commit into
mainfrom
AmesingFlank/stack/51

Commits