search

spcl / open-earth-compiler

development repository for the open earth compiler
https://arxiv.org/abs/2005.13014
Other
72 stars 14 forks source link

[mlir]the ptx compilation flow #45

Open CONGCONGLEEE opened 2 years ago

CONGCONGLEEE commented 2 years ago

Recently I'm trying to run the model compiled by onnx-mlir on GPU. Here is my idea about the compilation flow: affine/scf/std Dialect -> gpu Dialect ->NVVM IR -> LLVM IR -> ptx assemble. I have no idea whether it will work. If not, could you please share me some of your points? It would be very nice to get some help with this since I am not familiar with this area myself, thanks!

gysit commented 2 years ago

I think this is a good question for https://discourse.llvm.org/c/mlir/31 (I have not been working on GPU compilation recently).

The lowering path you propose sounds similar to what we have done in the open earth compiler but their have been a lot of changes since then. Having a look at something like IREE (https://github.com/google/iree) that AFAIK has an end-to-end GPU compilation workflow could help as well.

CONGCONGLEEE commented 2 years ago

Ok, thank you very much!