changdong1687

[FEA] Add INT8 support for sm_120 CollectiveBuilder (TmaWarpSpecialized)

2

### Which component requires the feature? CUTLASS C++ ### Feature Request SM120 `TmaWarpSpecialized builder` supports for `int_8` Hello, Currently, I am trying to write an INT8 GEMM kernel for the...

Hello, I build the kernels in RTX5090 with cuda-12.8, and when I call the fucntion `cutlass_scaled_nvfp4_mm_sm120` with bias is not `None`, it gives me a `RuntimeError: Error Internal` I tryed:...