llama.cpp
CUDA: bitonic argsort handles rows wider than one block
#28957

Merged

CUDA: bitonic argsort handles rows wider than one block #28957

IMbackK merged 1 commit into ggml-org:master from ServeurpersoCom:cuda-argsort-bitonic-wide-rows

ServeurpersoCom requested a review 21 days ago

github-actions added ggml

github-actions added CUDA

IMbackK assigned

IMbackK 21 days ago

JohannesGaessler approved these changes on 2026-09-17

CUDA: bitonic argsort handles rows wider than one block

f8fecfff

ServeurpersoCom force pushed from e3018985 to f8fecfff 13 days ago

IMbackK approved these changes on 2026-09-29

IMbackK merged 6a2743f0 into master 7 days ago

Reviewers

IMbackK

JohannesGaessler

Assignees

IMbackK

Labels

ggml CUDA

Milestone

No milestone