FFTc: An MLIR Dialect for Developing HPC Fast Fourier Transform Libraries

TL;DR

This paper introduces FFTc, a new MLIR-based domain-specific language designed to improve the development of high-performance Fourier Transform libraries for scientific computing.

Contribution

It presents the design, implementation, and initial results of FFTc, a novel MLIR dialect tailored for HPC Fourier Transform algorithms, inspired by FFTW.

Findings

Preliminary results show promising performance improvements.

Faster development of Fourier Transform libraries using FFTc.

Enhanced expressiveness for FFT algorithms in MLIR.

Abstract

Discrete Fourier Transform (DFT) libraries are one of the most critical software components for scientific computing. Inspired by FFTW, a widely used library for DFT HPC calculations, we apply compiler technologies for the development of HPC Fourier transform libraries. In this work, we introduce FFTc, a domain-specific language, based on Multi-Level Intermediate Representation (MLIR), for expressing Fourier Transform algorithms. We present the initial design, implementation, and preliminary results of FFTc.