llama.cpp
CUDA: Check PTX version on host side to guard PDL dispatch
#23530

Merged

CUDA: Check PTX version on host side to guard PDL dispatch #23530

JohannesGaessler merged 6 commits into ggml-org:master from ORippler:osimons/ptx_pdl_checks_cuda

CUDA: Check PTX version on host side to guard PDL dispatch

bcde3518

ORippler requested a review 42 days ago

ORippler requested a review from

JohannesGaessler 42 days ago

github-actions added Nvidia GPU

github-actions added ggml

JohannesGaessler commented on 2026-05-24

Implement MurmurHash3 mixer for better hash distribution

31449369

ORippler requested a review from

JohannesGaessler 37 days ago

JohannesGaessler commented on 2026-05-27

Update ggml/src/ggml-cuda/common.cuh

2c6d1f36

Address review comments, make seed non-zero

dc686584

Apply code-formatting

aaff9c2c

ORippler requested a review from

JohannesGaessler 37 days ago

JohannesGaessler approved these changes on 2026-05-27

Replace std::size_t -> size_t for consistency

864bac0d

ORippler requested a review from

gaugarg-nv 36 days ago

gaugarg-nv approved these changes on 2026-05-29

JohannesGaessler merged 6ed481ee into master 35 days ago

ORippler deleted the osimons/ptx_pdl_checks_cuda branch 35 days ago

Reviewers

gaugarg-nv

JohannesGaessler

Assignees

No one assigned

Labels

Nvidia GPU ggml

Milestone

No milestone