]> git.djapps.eu Git - pkg/ggml/sources/ggml/commit
overview / pkg / ggml / sources / ggml / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 0b157dc)
opencl: optimize mxfp4 kernels (llama/16037)
authorShawn Gu <redacted>
Thu, 18 Sep 2025 19:03:34 +0000 (12:03 -0700)
committerGeorgi Gerganov <redacted>
Sat, 20 Sep 2025 10:33:50 +0000 (13:33 +0300)
commitb7daf67c3b137e710fbfcf69eb771c629c1bf47d
treeea564e9589616288028ac21d54fce6a198cbb6d1tree
parent0b157dc701d1a2bac38ee5188265f5023896651acommit | diff
opencl: optimize mxfp4 kernels (llama/16037)

- flatten mxfp4 and packed fp4->fp16 bit-wise convert function (replace lut)
- MoE kernel optimizations

---------

Co-authored-by: Li He <redacted>
src/ggml-opencl/CMakeLists.txt diff | blob | history
src/ggml-opencl/ggml-opencl.cpp diff | blob | history
src/ggml-opencl/kernels/cvt.cl diff | blob | history
src/ggml-opencl/kernels/mul_mv_id_mxfp4_f32_flat.cl [new file with mode: 0644] blob
src/ggml-opencl/kernels/mul_mv_mxfp4_f32_flat.cl [new file with mode: 0644] blob
Packaging of ggml-org/ggml