llama.cpp
add avx2 for dot_q8_0_q8_0, 2x faster than scalar
#1211

Merged

add avx2 for dot_q8_0_q8_0, 2x faster than scalar #1211

sw merged 1 commit into ggml-org:master from YannFollet:dot_q8_0_q8_0_avx2

add avx2 for dot_q8_0_q8_0, 2x faster than scalar

e309138d

sw approved these changes on 2023-04-28

sw merged 04aaae1d into master 2 years ago

YannFollet deleted the dot_q8_0_q8_0_avx2 branch 1 year ago

Reviewers

Assignees

No one assigned

Labels

None yet

Milestone

No milestone