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

> Retained reddit record from reddit:LocalLLaMA.

Canonical: https://highsignal.app/data/records/4e5f712c682a9f0e

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:LocalLLaMA)  published 2026-09-26 06:18 ingested 2026-09-27 02:40

[original source ↗](https://www.reddit.com/r/LocalLLaMA/comments/1wqj4zg/ggmlcpu_tiled_mul_mat_for_kquants_by_jbooth_pull/)

## 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
