llama.cpp
ggml-cpu: Faster IQ1 mul_mat_vec on AVX2 using BMI2 instructions
#12154
Merged
Overview
Commits
6
Changes
View On GitHub

ggml-cpu: Faster IQ1 mul_mat_vec on AVX2 using BMI2 instructions #12154

slaren merged 6 commits into ggml-org:master from remyoudompheng:optim-x86
remyoudompheng
github-actions github-actions added ggml
slaren
JohnLoveJoy
remyoudompheng remyoudompheng force pushed from dd8f10c6 to d1aeed06 247 days ago
remyoudompheng
remyoudompheng ggml-cpu: Faster IQ1 mul_mat_vec on AVX2 using BMI2 instructions
91d7801f
remyoudompheng cmake: Add GGML_BMI2 build option
6feca684
remyoudompheng ggml: enable BMI2 on relevant CPU variants
3bfd4433
remyoudompheng ggml-cpu: include BMI2 in backend score
a3db5756
remyoudompheng remyoudompheng force pushed from 071c3129 to a3db5756 247 days ago
slaren
remyoudompheng ggml-cpu: register BMI2 in ggml_backend_cpu_get_features
1fc84888
remyoudompheng ggml-cpu: add __BMI2__ define when using MSVC
9d1281ab
remyoudompheng
slaren
slaren
slaren approved these changes on 2025-03-06
slaren slaren merged 07d15723 into master 245 days ago
remyoudompheng remyoudompheng deleted the optim-x86 branch 245 days ago
sandrohanea
sandrohanea
rudiservo
rudiservo

Login to write a write a comment.

Login via GitHub

Reviewers
Assignees
No one assigned
Labels
ggml
Milestone
No milestone
Terms Of Use Privacy Policy Refund Policy Impressum