| commit | 26e4c6b225a30466f93935e9d6d72b7a6d3f8155 | [log] [tgz] |
|---|---|---|
| author | Quinn Dawkins <quinn.dawkins@gmail.com> | Tue May 28 19:29:23 2024 -0400 |
| committer | GitHub <noreply@github.com> | Tue May 28 23:29:23 2024 +0000 |
| tree | 205a12e707c8bfeb17f71a2b6d1f083865cacb9e | |
| parent | ce1be9cd68b3e02d6a1a9240ccc19219f4c7614c [diff] |
[Codegen][GPU] Enable vectorization of iree_gpu.shuffle_tensor + lowering to iree_gpu barriers (#17506) This enables vectorization and lowering of iree_gpu.shuffle_tensor by adding a pattern to vectorize static results with the expectation that it connects with vectorization of adjacent operations. Additionally changes the lowering of iree_gpu.shuffle_tensor to generate iree_gpu barrier ops to keep things as a part of the SSA chain to improve analyzability and guarantee synchronization in the right places. Moves the shuffle_tensor lowering to the iree_gpu dialect because it makes more sense there.
IREE (Intermediate Representation Execution Environment, pronounced as “eerie”) is an MLIR-based end-to-end compiler and runtime that lowers Machine Learning (ML) models to a unified IR that scales up to meet the needs of the datacenter and down to satisfy the constraints and special considerations of mobile and edge deployments.
See our website for project details, user guides, and instructions on building from source.
IREE is still in its early phase. We have settled down on the overarching infrastructure and are actively improving various software components as well as project logistics. It is still quite far from ready for everyday use and is made available without any support at the moment. With that said, we welcome any kind of feedback on any communication channels!
See our website for more information.
Community meeting recordings: IREE YouTube channel
IREE is licensed under the terms of the Apache 2.0 License with LLVM Exceptions. See LICENSE for more information.