All pull requests

Pull requests

• Fix SM120 split-K alignment after partition caps

  #475·Sunt-ing opened Oct 15, 2026·

  ·
  1

• Add native FP8 Tensor Core GEMM support for SM89 (Ada Lovelace)

  #474·Ammar-Alnagar opened Oct 14, 2026·

  ·

• Add optimized fused FP8 x MXFP4 MegaMoE for SM90

  #473·AIExplorer7 opened Oct 14, 2026·

  ·

• docs: add Simplified Chinese translation (README.zh.md)

  #472·JasonYeYuhe opened Oct 14, 2026·

  ·

• Fix cross-expert output overwrites in SM90 masked GEMM

  #469·Sunt-ing opened Oct 12, 2026·

  ·

• Make pack_ue8m0_to_int CUDA graph capture safe

  #466·zhoubiao188 opened Oct 6, 2026·

  ·
  1

  1

• Update copyright year in LICENSE file

  #465·bishalramdam opened Oct 6, 2026·

  ·

• Update vendored CUTLASS to v4.8.0

  #464·fhgffy opened Oct 4, 2026·

  ·
  2

• Add SM90 BF16 grouped GEMM with gathered activation rows

  #463·yuweih205 opened Sep 30, 2026·

  ·

• [Fix] Bound cuBLASLt descriptor lifetime on errors

  #460·ZenAlexa opened Sep 28, 2026·

  ·

• [Fix] Honor deterministic algorithms in SM90 bmk,bnk->mn einsum

  #457·LiRunGuo opened Sep 22, 2026·

  ·
  4

• [Fix] Reject FP32 bmk,bnk->mn einsum without C instead of crashing

  #456·LiRunGuo opened Sep 22, 2026·

  ·
  2

• [Fix] SM90 heuristic offers block M 16/32 to the FP8 1D1D kernel (JIT failure for M <= 32)

  #455·LiRunGuo opened Sep 22, 2026·

  ·
  7

• Add SM90 NVFP4+FP8 MegaMoE support

  #454·lengrongfu opened Sep 21, 2026·

  ·
  24

• Fix SM120 BF16 GEMM shared-memory reuse ordering

  #453·Sunt-ing opened Sep 21, 2026·

  ·
  16

• perf: Use full-rate FADD relu in SM90 FP8 MQA logits epilogue

  #452·Chtholly-Boss opened Sep 18, 2026·

  ·
  4

• [Bugfix] Reset MegaMoE combine readiness for CUDA graph replay

  #451·foreverrookie opened Sep 17, 2026·

  ·
  3

• feat(gemm): opt-in labels-contract checker for m-grouped contiguous GEMM

  #450·lucifer1004 opened Sep 16, 2026Collaborator·

  ·
  18

• [BugFix][sm90] tf32 gemm race condition

  #448·leevan opened Sep 15, 2026·

  ·
  6

• Fix test helper discovery in pytest and sanitizer runner

  #446·lucifer1004 opened Sep 14, 2026Collaborator·

  ·
  8

• fix(sm90): support parameterized SwiGLU in FP8 MegaMoE

  #445·shiyang814-cpu opened Sep 14, 2026·

  ·
  11

• Add torch dispatcher op for tf32_hc_prenorm_gemm

  #442·leevan opened Sep 14, 2026·

  ·
  10

• Add a fused SM90 FP8 MegaMoE: one kernel, a lag-ordered schedule and a ring activation pool

  #440·zkyue opened Sep 13, 2026·

  ·
  15

• Fix unknown CMake command "cuda_add_library" in CMakeLists.txt

  #437·PerryLink opened Sep 10, 2026·

  ·
  4

• Improve SM100 MX MQA logits clean-output performance

  #436·OutWhite opened Sep 10, 2026·

  ·
  13