Issues

• Please provide a WHL for Windows

  JeffreyXiang/FlexGEMM#33 · Filter by author SoftologyPro opened on Oct 1, 2026

• CUDA driver error: device not ready on RTX 50-series (Blackwell, sm_120)

  JeffreyXiang/FlexGEMM#32 · Filter by author SaikatRoy17 opened on Sep 11, 2026

• Feature Proposal: CPU fallbacks for CUDA kernels (low VRAM support)

  JeffreyXiang/FlexGEMM#30 · Filter by author RisingRedstone opened on Jul 19, 2026

• [dev/all_triton branch] Inconsistent autotune cache keys for kernels wrapped with @triton.heuristics

  JeffreyXiang/FlexGEMM#26 · Filter by author kooWZ opened on May 9, 2026

• FlexGEMM 2.0：改进计划 (Roadmap)

  JeffreyXiang/FlexGEMM#23 · Filter by author EasternJournalist opened on May 8, 2026

• Build error at sparse_neighbor_map.cu

  JeffreyXiang/FlexGEMM#19 · Filter by author iiiCpu opened on Apr 14, 2026
  1

• Multi-GPU training crashes with flex_gemm

  JeffreyXiang/FlexGEMM#14 · Filter by author little-qjw opened on Feb 3, 2026

• Silent failure in RTX 5080

  JeffreyXiang/FlexGEMM#13 · Filter by author betaupsx86 opened on Jan 30, 2026

• cuda 13.0 support

  JeffreyXiang/FlexGEMM#8 · Filter by author bebelllm opened on Jan 13, 2026
  2