]> git.djapps.eu Git - pkg/ggml/sources/whisper.cpp/commit
vulkan: use vec dot for matrix matrix multiplications (llama/16056)
authorRuben Ortlam <redacted>
Sat, 20 Sep 2025 08:42:56 +0000 (10:42 +0200)
committerGeorgi Gerganov <redacted>
Sat, 20 Sep 2025 10:46:39 +0000 (13:46 +0300)
commit76d093428777161ae2759ef05a07c548269c3545
treeee7bbafaa696ef113375cf509f3141ff54b2f304
parent2ad00d558610608e1d9f4d2f217c4eb396630c90
vulkan: use vec dot for matrix matrix multiplications (llama/16056)

* vulkan: Change the mul_mm shared memory and register caching system to use vec2 instead of scalars, to enable using dot2 instructions

* use fma instead of dot to fix Nvidia and Apple performance issues
ggml/src/ggml-vulkan/vulkan-shaders/mul_mm.comp
ggml/src/ggml-vulkan/vulkan-shaders/mul_mm_funcs.comp
ggml/src/ggml-vulkan/vulkan-shaders/types.comp
ggml/src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp