Repository navigation
ggml-cuda: HIP: use __builtin_amdgcn_perm for q1_0 unpacking - #29927
Merged
JohannesGaessler merged 1 commit intoOct 17, 2026
Merged
JohannesGaessler merged 1 commit into
JohannesGaessler merged 1 commit into
Conversation
github-actions
Bot
added
ggml
Contributor
Author
|
Marked as DRAFT as i already have too many open PR, but it's ready for a review. @IMbackK |
This was referenced Oct 5, 2026
thelittlefireman
marked this pull request as ready for review
October 8, 2026 21:35
Contributor
|
@ggerganov this developer is making good contributions. I would like to retain them and have been thinking about inviting them to take a more active role in the project anyways. Can you make them a member? For context, their earliest PR was Aug 28. |
Member
I just sent them an invite to the @ggml-org/ggml-cuda team. |
thelittlefireman
force-pushed
the
optimize-hip-q1-amdgcn-perm
branch
from
October 15, 2026 20:02
b6c4ff7 to
efb7318
Compare
am17an
approved these changes
Oct 16, 2026
thelittlefireman
force-pushed
the
optimize-hip-q1-amdgcn-perm
branch
from
October 17, 2026 13:29
efb7318 to
d6f0bd9
Compare
Contributor
Author
|
@JohannesGaessler @IMbackK Can we merge this one ? |
JohannesGaessler
approved these changes
Oct 17, 2026
edwardyoon
pushed a commit
to edwardyoon/focus-llama
that referenced
this pull request
Oct 24, 2026
auRose94
pushed a commit
to auRose94/llama.cpp
that referenced
this pull request
Oct 28, 2026
1 task done
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Overview
Clean version of #28398. This PR handle both function
vec_dot_q1_0_q8_1andggml_cuda_mmq_load_tiles_q1_0by replacing__byte_permby__builtin_amdgcn_permfor performance.Additional information
AMD Instinct MI50 (gfx906),
m=4096,k=14336.Requirements