llama.cpp
hexagon: optimization for HMX mat_mul
#21554
Merged
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Overview
Commits
9
Changes
View On
GitHub
hexagon: optimization for HMX mat_mul
#21554
max-krasnyansky
merged 9 commits into
ggml-org:master
from
aizip:feat/hmx-optimization
njsyw1997
requested a review
72 days ago
github-actions
added
ggml
github-actions
added
Hexagon
njsyw1997
force pushed
from
f4b1704d
to
0aedac28
68 days ago
njsyw1997
force pushed
from
0aedac28
to
a51d65b2
68 days ago
njsyw1997
force pushed
from
0895350b
to
0d799771
68 days ago
max-krasnyansky
approved these changes on 2026-04-12
hexagon: add async HMX worker
de152555
hexagon: cost-based VTCM chunk search for out-stationary matmul
b2ec80a6
hexagon: fix futex race in hmx_worker_drain
f14e9c6f
hex-mm: hmx optimize scatter/transpose and use HMX intrinsics
f8737609
hex-vmem: drop vmem limit a touch under 3GB on v73
189a2d2f
hexagon: add fwd declaration of htp_context
6d82f7fb
hex-hmx: replace hmx-worker with hmx-queue that mimics dma-queue inteā¦
2af9a1c0
hex-mm: add debug log to hmx work func called from hmx-queue
c2b48b8f
njsyw1997
force pushed
from
0d799771
to
c2b48b8f
65 days ago
max-krasnyansky
approved these changes on 2026-04-14
max-krasnyansky
commented on 2026-04-14
Update hmx-queue.h
b122c95d
CISC
approved these changes on 2026-04-14
max-krasnyansky
merged
5d14e5d1
into master
65 days ago
Login to write a write a comment.
Login via GitHub
Reviewers
CISC
max-krasnyansky
Assignees
No one assigned
Labels
ggml
Hexagon
Milestone
No milestone
Login to write a write a comment.
Login via GitHub