ziyuhuang123

Emmmm, so the output is the same, right? I mean, outputs are just "tensor" type, no matter how I get a "tensor", I can do "copy(tCrC, tCsC);" and cute will...

Wow! That's cool! Actually I am writing relavant code pushed by my supervisor, haha.... Can not wait too long~He is a nice guy, I mean, anyway I am also interested...

Currently I only see one example code.... Do you possibly know more codes written by cute?

Oh, I noticed most gemm codes are written using previous cutlass.... Maybe I should use older version...? Because that doc is more....

And what is step??? How we use it? Any example or doc? Thanks!!

I see in another code, seems no need to use this Step? ``` Tensor gA = local_tile(A, make_tile(Int{}, Int{}), make_coord(iy, _)); Tensor gB = local_tile(B, make_tile(Int{}, Int{}), make_coord(ix, _)); Tensor...

I guess, just guess, if we have a tile 32, 32, 16, maybe we will first compute 32, 16, 16, and then compute the other 32, 16, 16, so the...