Fix gfx1103 performance regression#21720
Open
matteoserva wants to merge 2 commits intoggml-org:masterfrom
Open
Fix gfx1103 performance regression#21720matteoserva wants to merge 2 commits intoggml-org:masterfrom
matteoserva wants to merge 2 commits intoggml-org:masterfrom
Conversation
Commit 617db24 added nwarps=8 tuning for RDNA3_0 discrete GPUs (W7900/gfx1100), which causes ~30% throughput regression on gfx1103 iGPU (Radeon 780M/760M): 18 tok/s down to 13 tok/s. The iGPU has significantly less memory bandwidth and fewer CUs than discrete RDNA3 GPUs, so the higher warp count causes thread contention rather than improving throughput. Fix: detect gfx1103 specifically and use RDNA2 MMVQ parameters (nwarps=1) instead of RDNA3_0 parameters (nwarps=8). This restores pre-617db241 performance on iGPU while keeping the optimization for discrete RDNA3 GPUs. Changes: - vendors/hip.h: Add RDNA3_IGPU define for __gfx1103__ - common.cuh: Add GGML_CUDA_CC_RDNA3_IGPU constant and macro - mmvq.cu: Route gfx1103 to RDNA2 parameter table (nwarps=1) Fixes ggml-org#20647 Signed-off-by: Steve Westers <westers@users.noreply.github.com> Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
github-actions
bot
added
Nvidia GPU
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
Fixes the performance regression on gfx1103 (AMD IGPU used on mini pc)
Solves #20647
caused by 617db24
Requirements