vulkan: add v_dot2_f32_f16 support in matrix-matrix multiplication and Flash Attention#24123
Open
0cc4m wants to merge 4 commits into
Open
vulkan: add v_dot2_f32_f16 support in matrix-matrix multiplication and Flash Attention#241230cc4m wants to merge 4 commits into
v_dot2_f32_f16 support in matrix-matrix multiplication and Flash Attention#241230cc4m wants to merge 4 commits into
Conversation
jeffbolznv
reviewed
Jun 4, 2026
jeffbolznv
approved these changes
Jun 5, 2026
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
This PR adds basic support for the Vulkan extension
VK_VALVE_shader_mixed_float_dot_product. The background to this is that AMD Vega20, Navi14 and RDNA2+ GPUs have fp16 dot2 instructions for machine learning acceleration that are not emitted by the shader compiler due to numerical inconsistencies. The extension allows shaders to manually emit them.This PR adds support for the
v_dot2_f32_f16fp16 packed dot product with fp32 accumulator in matrix-matrix multiplications and Flash Attention. This is a good improvement for AMD GPUs with this instruction, but without coopmat support.AMD Radeon Pro VII (Vega20) Benchmarks
Requirements