linalg/block_quant: W4A8 GEMM for Q4_0 (int8-dot decode path)#2431
Open
czoli1976 wants to merge 1 commit into
Open
linalg/block_quant: W4A8 GEMM for Q4_0 (int8-dot decode path)#2431czoli1976 wants to merge 1 commit into
czoli1976 wants to merge 1 commit into
Conversation
Generalize the M==1 W4A8 GEMV to a GEMM that decodes each Q4_0 weight block once and reuses it across the M activation rows, computing per-block int8 dot products of the unpacked 4-bit weights against int8-quantized activations. Writes into a caller-provided [m, n] buffer to stay allocation-free on the hot path. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Follow-up to #2348: generalises the
M == 1W4A8 GEMV to a GEMM that decodes each Q4_0 weight block once and reuses it across theMactivation rows, writing into a caller-provided[m, n]buffer — the allocation-free&mut [f32]output you flagged on #2348. Correctness is covered by asserting the GEMM is bit-identical to the per-row GEMV, plus an approximate check against the f32-dequant matmul.🍍