Skip to content

Commit

Permalink
Vulkan: VK_KHR_cooperative_matrix support to speed up prompt processi…
Browse files Browse the repository at this point in the history
…ng (#10597)

* Vulkan: Implement VK_KHR_cooperative_matrix support in the matrix matrix multiplication shader

* Improve performance with better q4_k and q5_k dequant and store unrolling

* Add Vulkan MUL_MAT and MUL_MAT_ID accumulator precision selection

* Rework mulmat shader selection and compilation logic, avoid compiling shaders that won't get used by device

* Vulkan: Implement accumulator switch for specific mul mat mat shaders

* Vulkan: Unroll more loops for more mul mat mat performance

* Vulkan: Add VK_AMD_shader_core_properties2 support to read Compute Unit count for split_k logic

* Disable coopmat support on AMD proprietary driver

* Remove redundant checks

* Add environment variable GGML_VK_DISABLE_COOPMAT to disable VK_KHR_cooperative_matrix support

* Fix rebase typo

* Fix coopmat2 MUL_MAT_ID pipeline selection
  • Loading branch information
0cc4m authored Dec 7, 2024
1 parent 86a1934 commit 3df784b
Show file tree
Hide file tree
Showing 3 changed files with 750 additions and 397 deletions.
Loading

0 comments on commit 3df784b

Please sign in to comment.