llama.cpp
Workaround for subgroup arithmetic failing on MoltenVK with AMD GPUs
#15886
Merged
Merged
Workaround for subgroup arithmetic failing on MoltenVK with AMD GPUs #15886
0cc4m
merged
1 commit into
ggml-org:master
from
lksj92hs:lksj92hs/moltenvk-subgroup-workaround
Workaround for subgroup arithmetic failing on MoltenVK with AMD GPUs …
8d6c2ea9
0cc4m
approved these changes
on 2025-09-09
Login to write a write a comment.
Login via GitHub