PR #26340 [QNN-EP] Add MatMulNBits translation for GPU

[QNN-EP] Add MatMulNBits translation for GPU #26340

yuslepukhin merged 10 commits into microsoft:main from CodeLinaro:dev/tirupath/matmulnbits_gpu

jambayk commented on 2025-10-17

edgchen1 commented on 2025-10-17

edgchen1 commented on 2025-10-20

johnpaultaken approved these changes on 2025-11-19

edgchen1 commented on 2025-11-20

edgchen1 added ep:QNN

quic-tirupath force pushed from e83f6e50 to b654efea 260 days ago

edgchen1 commented on 2025-12-16

tirupath-qti force pushed from b654efea to c40029e5 255 days ago

edgchen1 commented on 2026-01-08

johnpaultaken commented on 2026-01-11

yuslepukhin requested a review from

copilot-pull-request-reviewer 229 days ago

copilot-pull-request-reviewer commented on 2026-01-12

tirupath-qti force pushed from c00dfbc1 to 241f8434 227 days ago

edgchen1 commented on 2026-01-14

tirupath-qti commented on 2026-01-15

edgchen1 dismissed these changes on 2026-01-15

edgchen1 added release:1.24.0

[QNN-EP] Add MatMulNBits translation for GPU

a59209bc

[QNN Ep] Fix minor issues in MatMulNBits translation for GPU

55467911

[QNN EP] Fix overflow and type casting issues

2d6ced3b

Fix variable name

a4d710cd

Fix bias input exist check in MatMulNBits op builder

3c045459

fix review comments

010ee596

Fix the copilot suggestion

df39d493

update unique_ptr to be of type uint32_t and Qnn_ScaleOffset_t

374aae2e

Add additional check on input configuration

9a8747f0

tirupath-qti force pushed from 6c3325e2 to 9a8747f0 226 days ago

Merge branch 'main' into dev/tirupath/matmulnbits_gpu

736855a2

tirupath-qti dismissed their stale review via 736855a2 226 days ago

yuslepukhin approved these changes on 2026-01-15

yuslepukhin enabled auto-merge (squash) 226 days ago

yuslepukhin merged ba11af41 into main 226 days ago

tianleiwu removed release:1.24.0

tianleiwu added cherry-picked

Reviewers

yuslepukhin

johnpaultaken

edgchen1

jambayk

tirupath-qti

copilot-pull-request-reviewer

skadaver-qti

Assignees

No one assigned

Labels

ep:QNN cherry-picked

Milestone

No milestone

onnxruntime [QNN-EP] Add MatMulNBits translation for GPU #26340 Merged

[QNN-EP] Add MatMulNBits translation for GPU #26340

onnxruntime
[QNN-EP] Add MatMulNBits translation for GPU
#26340

Merged