MLIR Projects Directory

This page aims to list all public MLIR projects at Intel, in this repository, separate forks or other projects that use LLVM (ex. standalone).

If your project isn't listed here, please add them directly or ask someone in the repo to add them for you.

Note: The formatting on this page is experimental. Feel free to play with it until we find a more efficient way of representing all projects.

Table Of Contents

1. IMEX
2. Region Op
3. SYCL MLIR
4. TPP MLIR
5. GENX

Projects

IMEX

Name	IMEX
Repo	https://github.com/intel/mlir-extensions
Blurb	Intel® Extension for MLIR. A staging ground for MLIR dialects and tools for Intel devices using the MLIR toolchain.
Main Contact	https://github.com/Jianhui-Li

Region Op

Name	region-op
Repo	https://github.com/intel/llvm/tree/region-op
Blurb	Prototyping a region operation to group code by device and/or memory configuration
Main Contact	https://github.com/Hardcode84, https://github.com/rengolin

SYCL-MLIR

Name	SYCL-MLIR
Repo	https://github.com/intel/llvm/tree/sycl-mlir
Blurb	An MLIR based optimizing compiler for the SYCL programming language.
Main Contact	https://github.com/etiotto, https://github.com/whitneywhtsang

TPP MLIR

Name	TPP MLIR
Repo	https://github.com/plaidml/tpp-mlir
Blurb	TPP experimentation on MLIR for linear algebra
Main Contact	https://github.com/rengolin
Description	This is an experiment in using MLIR to automatically select the best Tensor Processing Primitives for linear algebra.
Collaborates with	IMEX, OpenXLA/IREE, region-op

GENX

Name	GENX
Repo	https://github.com/intel/llvm/tree/genx
Blurb	GENX Dialect in MLIR (extends LLVM IR for Intel GPUs)
Main Contact	https://github.com/etiotto
Description	The GENX dialect is the counterpart of the NVVM/ROCDL MLIR dialects. This dialect extends the LLVM IR dialect with operations specific to Intel GPUs