]> git.djapps.eu Git - pkg/ggml/sources/ggml/commit
overview / pkg / ggml / sources / ggml / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: db51a43)
ggml : Leverage the existing GGML_F32_VEC helpers to vectorize ggml_vec_set_f32 for...
authorsirus20x6 <redacted>
Wed, 22 Oct 2025 10:14:14 +0000 (05:14 -0500)
committerGeorgi Gerganov <redacted>
Sat, 1 Nov 2025 07:41:35 +0000 (09:41 +0200)
commit3c8967a1ac3110ce2da83071d771fb1608430db0
treee36495959e55a0672616677c2ca6239428473911tree
parentdb51a43467eb7e0b410ce06265e9608eb16a267ccommit | diff
ggml : Leverage the existing GGML_F32_VEC helpers to vectorize ggml_vec_set_f32 for faster fills (llama/16522)

* Leverage the existing GGML_F32_VEC helpers to broadcast the fill value across SIMD registers and store in vector-sized chunks, while retaining the scalar tail for leftover elements and non-SIMD builds.

* Vectorize additional f32 helper loops

* Normalize f32 helper tails for ggml vec ops

---------

Co-authored-by: Aaron <redacted>
src/ggml-cpu/vec.h diff | blob | history
Packaging of ggml-org/ggml