CUDA Plugin Windows CI

Workflow runs for CUDA Plugin Windows CI

• [CUDA] Add fast INT3 QMoE kernels and offline SM80 INT2/INT4 prepack support

  currently running:
  CUDA Plugin Windows CI#12495:Pull request#33569synchronizebytianleiwu
  tlwu/20261023/qmoe-int3-int2-prepack-cuda

  Oct 24, 2026In progress
  More actions for workflow run 12495

• [CUDA] Optimize HyperConnectionLowRankMix 4-row decode, rank-320 specialization, and deterministic prefill

  currently running:
  CUDA Plugin Windows CI#12494:Pull request#33548synchronizebytianleiwu
  asonawane/dflash2-s4-hyperconnection

  Oct 24, 2026In progress
  More actions for workflow run 12494

• [CUDA] Extend DynamicSparseAttention INT8 KV cache and PackedSparseAttentionIndexer QSA

  failed:
  CUDA Plugin Windows CI#12493:Pull request#33472synchronizebytianleiwu
  asonawane/main-ops-sparse-attention-indexer

  Oct 24, 202650m 46s
  More actions for workflow run 12493

• [CUDA] Extend GatedDeltaNet and CausalConvWithState for S<=8 decode and prefix states

  currently running:
  CUDA Plugin Windows CI#12492:Pull request#33471synchronizebytianleiwu
  asonawane/main-ops-gdn-convolution

  Oct 24, 2026In progress
  More actions for workflow run 12492

• [CUDA] Add RMSNormRotary contrib operator

  currently running:
  CUDA Plugin Windows CI#12491:Pull request#33469synchronizebytianleiwu
  asonawane/main-ops-normalization-rotary

  Oct 24, 2026In progress
  More actions for workflow run 12491

• [CUDA] Harden GatedDeltaNet BFloat16 fp16-mode guards and add split-engine fp16 path

  queued:
  CUDA Plugin Windows CI#12490:Pull request#33523synchronizebytianleiwu
  tlwu/20261021/gdn-bf16-fp16-precision-fixes

  Oct 24, 2026Queued
  More actions for workflow run 12490

• [CUDA] Add session-scoped MoE/QMoE offload planner, lifecycle controller, and memory budget ledger

  currently running:
  CUDA Plugin Windows CI#12489:Pull request#33451synchronizebytianleiwu
  tlwu/20261019/qmoe-offload-p2-cuda-memory-lifecycle

  Oct 23, 2026In progress
  More actions for workflow run 12489

• [CUDA] Add FoldMid2 and FoldTail speculative state fold contrib operators

  failed:
  CUDA Plugin Windows CI#12488:Pull request#33474synchronizebytianleiwu
  asonawane/main-ops-mtp-fold-state

  Oct 23, 20261h 12m 23s
  More actions for workflow run 12488

• Test Node.js encrypted compiled-model inference

  currently running:
  CUDA Plugin Windows CI#12487:Pull request#33219synchronizebyGopalakrishnanN
  GopalakrishnanN/test-encrypted-epcontext-node

  Oct 23, 2026In progress
  More actions for workflow run 12487

• [Objective-C] Add encrypted compiled EPContext inference coverage

  queued:
  CUDA Plugin Windows CI#12486:Pull request#33223synchronizebyGopalakrishnanN
  GopalakrishnanN/test-encrypted-epcontext-objectivec

  Oct 23, 2026Queued
  More actions for workflow run 12486

• [CPU/MLAS] Add QMoE shared 4-bit expert layouts, AVX-512 kernels, and FileExpertStore

  completed successfully:
  CUDA Plugin Windows CI#12485:Pull request#33450synchronizebytianleiwu
  tlwu/20261019/qmoe-offload-p1-cpu-shared-layout

  Oct 23, 20261h 23m 29s
  More actions for workflow run 12485

• [CPU/MLAS] Add QMoE shared 4-bit expert layouts, AVX-512 kernels, and FileExpertStore

  cancelled:
  CUDA Plugin Windows CI#12484:Pull request#33450synchronizebytianleiwu
  tlwu/20261019/qmoe-offload-p1-cpu-shared-layout

  Oct 23, 20261m 47s
  More actions for workflow run 12484

• [CUDA] Fix Windows nvcc error for FP4 rescale device variable templat…

  currently running:
  CUDA Plugin Windows CI#12483:Commit36f07f7pushedbytianleiwu
  main

  Oct 23, 2026In progress
  More actions for workflow run 12483

• [CUDA] Extend DynamicSparseAttention INT8 KV cache and PackedSparseAttentionIndexer QSA

  cancelled:
  CUDA Plugin Windows CI#12482:Pull request#33472synchronizebytianleiwu
  asonawane/main-ops-sparse-attention-indexer

  Oct 23, 202655m 50s
  More actions for workflow run 12482

• [CUDA] Add SigmoidGateMul and ArgMaxLastAxis contrib operators

  queued:
  CUDA Plugin Windows CI#12481:Pull request#33470synchronizebytianleiwu
  asonawane/main-ops-hyperconnection-tensor

  Oct 23, 2026Queued
  More actions for workflow run 12481

• [CUDA] Add RMSNormRotary contrib operator

  cancelled:
  CUDA Plugin Windows CI#12480:Pull request#33469synchronizebytianleiwu
  asonawane/main-ops-normalization-rotary

  Oct 23, 202650m 17s
  More actions for workflow run 12480

• [CUDA] Add FoldMid2 and FoldTail speculative state fold contrib operators

  cancelled:
  CUDA Plugin Windows CI#12479:Pull request#33474synchronizebytianleiwu
  asonawane/main-ops-mtp-fold-state

  Oct 23, 202617m 57s
  More actions for workflow run 12479

• Retain external-data file handles and original model directories

  completed successfully:
  CUDA Plugin Windows CI#12478:Pull request#33062synchronizebyGopalakrishnanN
  GopalakrishnanN:gnallasamy-microsoft-external-data-file-handles

  Oct 23, 20261h 35m 32s
  More actions for workflow run 12478

• [CUDA] Extend DynamicSparseAttention INT8 KV cache and PackedSparseAttentionIndexer QSA

  cancelled:
  CUDA Plugin Windows CI#12477:Pull request#33472synchronizebytianleiwu
  asonawane/main-ops-sparse-attention-indexer

  Oct 23, 202613m 41s
  More actions for workflow run 12477

• Fix QMoE SM80 FP4 init order: don't let the tactic seed throw (#32194)

  completed successfully:
  CUDA Plugin Windows CI#12476:Commit5d466bapushedbyGopalakrishnanN
  main

  Oct 23, 20261h 34m 0s
  More actions for workflow run 12476

• [CUDA] Extend GatedDeltaNet and CausalConvWithState for S<=8 decode and prefix states

  cancelled:
  CUDA Plugin Windows CI#12475:Pull request#33471synchronizebytianleiwu
  asonawane/main-ops-gdn-convolution

  Oct 23, 20261h 11m 59s
  More actions for workflow run 12475

• [WebGPU] Batch streamless GPU-to-GPU copy vectors

  completed successfully:
  CUDA Plugin Windows CI#12474:Pull request#33567synchronizebyaciddelgado
  aciddelgado/fixed-state-batching

  Oct 23, 20261h 28m 47s
  More actions for workflow run 12474

• [CUDA] Add RMSNormRotary contrib operator

  cancelled:
  CUDA Plugin Windows CI#12473:Pull request#33469synchronizebytianleiwu
  asonawane/main-ops-normalization-rotary

  Oct 23, 202627m 38s
  More actions for workflow run 12473

• [CUDA] Add SigmoidGateMul and ArgMaxLastAxis contrib operators

  cancelled:
  CUDA Plugin Windows CI#12472:Pull request#33470synchronizebytianleiwu
  asonawane/main-ops-hyperconnection-tensor

  Oct 23, 202634m 36s
  More actions for workflow run 12472

• [WebGPU] Batch streamless GPU-to-GPU copy vectors

  cancelled:
  CUDA Plugin Windows CI#12471:Pull request#33567synchronizebyaciddelgado
  aciddelgado/fixed-state-batching

  Oct 23, 202629m 56s
  More actions for workflow run 12471