]> git.djapps.eu Git - pkg/ggml/sources/ggml/commit
HIP: RDNA4 tensor core support for MMF (llama/17077)
authoryulo <redacted>
Fri, 21 Nov 2025 23:03:24 +0000 (07:03 +0800)
committerGeorgi Gerganov <redacted>
Thu, 11 Dec 2025 13:32:42 +0000 (15:32 +0200)
commit499b79717b920f7040ad06f34b0e23784b737454
tree2d058280dcfacd59699a8f3471d86e7fe11e5833
parent774088bcc1c7207becf055038ac9602239ebeb7c
HIP: RDNA4 tensor core support for MMF (llama/17077)

* mmf for rdna4

* align the padding for rdna4

* forbit mul_mat_f for rdna4

* fix as comment

* remove device kernels

* add constexpr for early return

* update based on review comment

* change based on the review comment

* pass compile error

* keep code consistency

---------

Co-authored-by: zhang hui <redacted>
src/ggml-cuda/common.cuh
src/ggml-cuda/convert.cuh
src/ggml-cuda/mma.cuh
src/ggml-cuda/mmf.cu
src/ggml-cuda/mmf.cuh