← sources

ggml-cpu: tiled mul_mat for k-quants by jbooth · Pull Request #27851 · ggml-org/llama.cpp

source: reddit (reddit:LocalLLaMA)published 2026-09-26 06:18ingested 2026-09-27 02:40

retained text

faster CPU prompt processing: "TL;DR: 3-7x faster CPU mul\_mat using VNNI with IMO minimal complexity"



document fetched 2026-09-27 00:39 · canonical https://reddit.com/r/LocalLLaMA/comments/1wqj4zg/ggmlcpu_tiled_mul_mat_for_kquants_by_jbooth_pull