cutlass icon indicating copy to clipboard operation
cutlass copied to clipboard
verified publisher icon NVIDIA

[QST][CuteDSL] warp mma support

Open jinzhen-lin opened this issue 1 month ago • 0 comments

I have noticed that CuteDSL only supports fp16/bf16 warp mma with shape m16n8k16 and m16n8k8 now.

https://github.com/NVIDIA/cutlass/blob/ec8daf642d69fc31352ac6fa6e14a0de9019604b/python/CuTeDSL/cutlass/cute/nvgpu/warp/mma.py

Are there any other support plans in the future, such as

  • Turing support
  • Volta support (with mma shape m8n8k4)
  • B1 / INT4 / INT8 / FP4 / FP6 / FP8 / TF32 support

BTW, the documentation comments for MmaF16BF16Op seem to be incorrect. My understanding is that this is not tcgen05, right?

https://github.com/NVIDIA/cutlass/blob/ec8daf642d69fc31352ac6fa6e14a0de9019604b/python/CuTeDSL/cutlass/cute/nvgpu/warp/mma.py#L43-L50

jinzhen-lin avatar Dec 01 '25 04:12 jinzhen-lin