Skip to content

feat(qy): enable InfiniCCL NCCL backend for Denglin QY (2026.11.12) - #90

Open
Ceng23333 wants to merge 1 commit into
InfiniTensor:masterfrom
Ceng23333:feat/qy-nccl-20261112
Open

Ceng23333 wants to merge 1 commit into
InfiniTensor:masterfrom
Ceng23333:feat/qy-nccl-20261112

Conversation

@Ceng23333

@Ceng23333 Ceng23333 commented Nov 12, 2026 •

Copy link
Copy Markdown

Summary

Enable Denglin QY device + NCCL CclApi on current master so InfiniLM TP (GetUniqueId/CommInitRank/AllReduce/Send/Recv) works on QY.

Also adds WITH_QY CMake detection/SDK configure and NCCL backend gating for QY.

Validation

  • Cluster/example programs: not run on PR host (no QY SDK).
  • Integration-validated via ITW v2026.11.12-qy-moe-bf16-warpnk.

Stack PR index (merge order)

  1. InfiniOps: feat(qy): QY ops backend with FA and longctx MoE/RoPE kernels (2026.11.12) InfiniOps#1038
  2. InfiniRT: feat(qy): enable QY CUDA runtime device backend (2026.11.12) InfiniRT#54
  3. InfiniCCL: feat(qy): enable InfiniCCL NCCL backend for Denglin QY (2026.11.12) #90
  4. InfiniCore: chore(qy): pin InfiniRT/Ops/CCL for QY MoE bf16-warpnk (2026.11.12) InfiniCore#1583
  5. InfiniLM: feat(qy): MiniCPM5 FA/longctx/cudagraph + MoE bf16/warpnk surpass (2026.11.12) InfiniLM#618 → base refactor/adopt-modern-infini-stack

ITW pin: https://github.com/Ceng23333/InfiniTensorWorktree/tree/v2026.11.12-qy-moe-bf16-warpnk

Supersedes working tip of closed #77 (CPU header already on master; this PR adds QY NCCL).

Wire WITH_QY device detection/SDK configure, DevicePriority/NCCL
IsSupportedCombination for kQy, and QY device/runtime/NCCL CclApi so TP
GetUniqueId/CommInitRank/AllReduce/Send/Recv work on QY.

Co-authored-by: Cursor <cursoragent@cursor.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant