]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
cuda : add RDNA4-specific MMVQ parameter table for bs=1 decode (#19478)
authorPikaPikachu <redacted>
Sun, 15 Mar 2026 07:33:39 +0000 (15:33 +0800)
committerGitHub <redacted>
Sun, 15 Mar 2026 07:33:39 +0000 (08:33 +0100)
commit617db241aac17069ef43743b31ef1ac3105117aa
tree64a1b292a397d616dcdd9dacdc3cb189de92dd98
parent1a3d8edbbaba7f6e36096982c7c8a7ce11f4a7e8
cuda : add RDNA4-specific MMVQ parameter table for bs=1 decode (#19478)

* mmvq: add RDNA3/RDNA4-specific parameter table (nwarps=8, rows=1)

* mmvq: add dedicated RDNA3 parameter table

* mmvq: exclude RDNA3.5 (gfx1150/1151) from RDNA3 table
ggml/src/ggml-cuda/mmvq.cu
ggml/src/ggml-cuda/vendors/hip.h