llama.cpp
ggml-webgpu: performance-portable matmul tuning knobs
#22241

Merged

ggml-webgpu: performance-portable matmul tuning knobs #22241

reeselevine merged 1 commit into ggml-org:master from SharmaRithik:performance-portable-matmul-tuning-knobs

SharmaRithik requested a review 51 days ago

github-actions added ggml

github-actions added WebGPU

SharmaRithik force pushed from 0269022c to 0696db40 50 days ago

SharmaRithik force pushed from 0696db40 to 13112fd3 49 days ago

SharmaRithik force pushed from 13112fd3 to 99e94e31 49 days ago

reeselevine commented on 2026-04-24

add performance-portable tuning for register-tile and subgroup matmul

a059743d

SharmaRithik force pushed from 99e94e31 to a059743d 49 days ago

reeselevine approved these changes on 2026-04-24

reeselevine requested a review from

CISC 47 days ago

reeselevine requested a review from

ggerganov 47 days ago

CISC approved these changes on 2026-04-26

reeselevine merged 06a811d0 into master 47 days ago

Reviewers

CISC

reeselevine

ggerganov

Assignees

No one assigned

Labels

ggml WebGPU

Milestone

No milestone