llama.cpp
437e7785 - 10+% performance improvement of ggml_vec_dot_q4_0 on AVX2 (#654)

Commit

2 years ago

10+% performance improvement of ggml_vec_dot_q4_0 on AVX2 (#654) * Performance improvement of AVX2 code * Fixed problem with MSVC compiler * Reviewer comments: removed double semicolon, deleted empty line 1962

References

#654 - 10+% performance improvement of ggml_vec_dot_q4_0 on AVX2

Author

SebastianApel

Parents

cd7fa956

Files1

ggml.c

llama.cpp 437e7785 - 10+% performance improvement of ggml_vec_dot_q4_0 on AVX2 (#654)

llama.cpp
437e7785 - 10+% performance improvement of ggml_vec_dot_q4_0 on AVX2 (#654)