llama.cpp
Rotate K/V read order for better L2 cache hit rates
#32407
Merged

Rotate K/V read order for better L2 cache hit rates #32407

gaugarg-nv merged 1 commit into ggml-org:master from gaugarg-nv:opt_fattn
gaugarg-nv
gaugarg-nv Rotate K/V read order for better L2 cache hit rates
7004be97
gaugarg-nv gaugarg-nv requested a review 3 days ago
github-actions github-actions added ggml
github-actions github-actions added CUDA
am17an
am17an approved these changes on 2026-11-17
anujj
anujj approved these changes on 2026-11-17
gaugarg-nv gaugarg-nv merged 054a1bab into master 2 days ago
gaugarg-nv gaugarg-nv deleted the opt_fattn branch 2 days ago

Login to write a write a comment.

Login via GitHub

Reviewers
Assignees
No one assigned
Labels
Milestone