High Performance Code Generation in MLIR: An Early Case Study with GEMM

Abstract

This article is primarily meant to present an early case study on using MLIR, a new compiler intermediate representation infrastructure, for high-performance code generation. Aspects of MLIR covered in particular include memrefs, the affine dialect, and polyhedral utilities and pass infrastructure surrounding those. This article is also aimed at showing the role compiler infrastructure could play in generating code that is competitive with highly tuned manually developed libraries, albeit in a more modular, reusable, and automatable way.

References in corpus (1)

• MLIR: A Compiler Infrastructure for the End of Moore's Law

Cited by in corpus (3)

• HIR: An MLIR-based Intermediate Representation for Hardware Accelerator Description
• Autotuning PolyBench Benchmarks with LLVM Clang/Polly Loop Optimization Pragmas Using Bayesian Optimization (extended version)
• Autotuning PolyBench Benchmarks with LLVM Clang/Polly Loop Optimization Pragmas Using Bayesian Optimization