llama.cpp
ggml-cpu: add 128-bit RVV implementation for Quantization Vector Dot
#20633

Open

ggml-cpu: add 128-bit RVV implementation for Quantization Vector Dot #20633

rehan-10xengineer wants to merge 3 commits into ggml-org:master from riseproject-dev:10x/riscv-quant-vec-dot-128b

rehan-10xengineer requested a review from

ggerganov 17 days ago

github-actions added ggml

taimur-10x force pushed from c7c6abc3 to d618925f 17 days ago

ggerganov requested a review from

xctan 17 days ago

ggerganov requested a review from

copilot-pull-request-reviewer 17 days ago

copilot-pull-request-reviewer commented on 2026-03-16

xctan approved these changes on 2026-03-18

ggml-cpu: add 128-bit impls for i-quants, ternary quants

2fe760f0

ggml-cpu: add 128-bit impls for iq2_xs, iq3_s, iq3_xxs, tq2_0

4b12d409

taimur-10x force pushed from d618925f to cf95828a 15 days ago

ggml-cpu: refactor; add rvv checks

05a5425e

taimur-10x force pushed from cf95828a to 05a5425e 15 days ago

Reviewers

xctan

ggerganov

Assignees

No one assigned

Labels

ggml

Milestone

No milestone