Sync with Microsoft ONNX Runtime - 24052026#1104

Closed

ai-fw-intg wants to merge 17 commits into

ovep-developfrom

sync_msft_24052026

Commits on May 21, 2026

Fix oob bias access for MatMulIntegerToFloat and DynamicQuantizeMatMul (#28499)
adrianlizarraga
authored
Validate conv bias shape in WordConvEmbedding to prevent OOB read (#28279)
apsonawane
authored
[CoreML EP] Support Gather with scalar 'indices' (#28278)

maxwbuckley
and
claude
authored

Commits on May 22, 2026

Add component governance manifest for WebGPU EP (#28599)

adrastogi
and
Aditya Rastogi
authored
Validate seqlens_k against cos_cache bounds in GroupQueryAttention to… (#28277)
apsonawane
authored
Parallelize CPU ScatterElements kernel via ThreadPool (#28588)

Copilot
and
tianleiwu
authored
[NVEP]: fix test for multi-gpu situation (#27837)
theHamsta
authored
QMoE CUDA: input validation, prepack cleanups, and packaging pipeline fix (#28607)
tianleiwu
authored
Fix CUDA build with contrib ops disabled (#28554)

authored
Optimize MLAS quantized KV-cache GEMM kernels (follow-up to #28578) (#28606)
tianleiwu
authored
fix(qdq): skip DQ forward propagation when DQ input is constant (#28521)
Rishi-Dave
authored
[WebGPU] LinearAttention: increase tile_v when subgroups are available (#28519)
daijh
authored
Fix Reshape with allowzero=1 producing wrong shape for zero-size tensor in chained Reshape (#28455)

Copilot
and
tianleiwu
authored
Fix: Accept 'CPU' as a valid provider name in SessionOptionsAppendExecutionProvider (#28625)

jiafatom
and
Copilot
authored

Commits on May 23, 2026