et-backend: Q8_0 matrix-engine GEMM kernel for prefill#14
Closed
RehanQasim-dev wants to merge 1 commit into
Closed
et-backend: Q8_0 matrix-engine GEMM kernel for prefill#14RehanQasim-dev wants to merge 1 commit into
RehanQasim-dev wants to merge 1 commit into
Conversation
Adds a Q8_0 matrix-engine mul_mat kernel with static-reuse dequant across N-tiles and K-splitting, plus fused bias support for MUL_MAT+ADD. Routes N >= 47 to the matrix engine and N <= 47 to vecdot. Registers the new kernel in the KERNELS list, which it was missing from, so the runtime kernel lookup was silently failing and leaving the destination buffer unwritten.
Author
|
Superseded by the upstream PR against ggml-org/llama.cpp. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Performance (Llama-3.2-1B-Instruct, ET-SoC-1)
Prefill improved roughly 6-9x. Decode is the same as et since decode (N=1) always routes to the vecdot kernel, not the matrix-engine path this PR touches.
Prefill t/s
Testing
Verified with llama-cli on ET-SoC-1 hardware (Llama-3.2-1B-Instruct), coherent output.