[QNN EP] 2.3.0 RC2 Changes#423

Merged

qti-mbadnara merged 12 commits into

dev/qti-mbadnara/rel-2.3.0-rc2-changes

May 21, 2026

Commits on May 20, 2026

[QNN EP] Ensure ORT Core is 1.24.4 (#369)
qti-mbadnara
committed
Re-enable failed UT for Linux aarch64 running on qcs6490 (#361)

chunghow-qti
authored and
qti-mbadnara
committed
Remove unnecessary content in ThirdPartyNotice (#351)

qti-hungjuiw
authored and
qti-mbadnara
committed
[QNN EP] Fuse Dynamic MatMulInteger pattern into Float QNN MatMul (#367)

tirupath-qti
authored and
qti-mbadnara
committed
[QNN EP] Uplevel QAIRT to 2.46.0 (#370)
qti-mbadnara
committed
[QNN EP] Fix ReshapeGemmFusion rank-5 input regression (#377)

huaychou
authored and
qti-mbadnara
committed
[QNN EP] GemmOpBuilder: add support for beta=0.0 (#375)

huaychou
authored and
qti-mbadnara
committed
[QNN EP] Fuse DynamicQuantizeLinear + ConvInteger pattern (#364)

tirupath-qti
authored and
qti-mbadnara
committed
[QNN EP] Minor CI Fixes (#395)
qti-mbadnara
committed
[QNN EP] Add support to build ARM64 (ARM64x) Zip (#397)
qti-mbadnara
committed
[QNN EP] Support coordinate_transformation_mode=half_pixel and sampling_ratio=0 in RoiAlign (#389)

huaychou
authored and
qti-mbadnara
committed
[OpBuilder] Support HTP MatMulNBits. (#288)

minfhong-qti
authored and
qti-mbadnara
committed