FFTX and SPIRAL

(DE-AC02-05CH11231 — DOE FFTX)

FFTX is the exascale follow-on to the FFTW open source discrete FFT package for executing the Fast Fourier Transform as well as higher-level operations composed of linear operations combined with DFT transforms. Though backwards compatible with FFTW, this is an entirely new work developed as a cooperative effort between Lawrence Berkeley National Laboratory, Carnegie Mellon University, and SpiralGen, Inc.

From casual look at its API, FFTX appears to be a pre-built library, but the heart of FFTX is a build-time code generator, SPIRAL, that produces very high performance kernels targeted to their specific uses and platform environments. Coupled to the platform-aware code generator is a sophisticated front end that interprets the details of the algorithms from the FFTX API, which it treats as a DSL for algorithm specification.

 

Development

Download

FFTX release 1.0.0 is available on GitHub, along with the supporting releases of Spiral and Spiral packages. All are open source under a BSD-style license.

• FFTX
• spiral-software
• spiral-package-fftx
• spiral-package-simt
• spiral-package-mpi

References

• F. Franchetti, D. G. Spampinato, A. Kulkarni, T. M. Low, M. Franusich, D. T. Popovici, A. Canning, P. McCorquodale, B. Van Straalen, P. Collela
  FFT and Solver Libraries for Exascale: FFTX and SpectralPack
  Exascale Computing Project (ECP) Annual Meeting 2020, Poster


• A. Kulkarni, D. Spampinato, F. Franchetti
  FFTX for Micromechanical Stress-Strain Analysis
  IEEE High Performance Extreme Computing Conference (HPEC) 2019, Poster with extended abstract


• A. Kulkarni, D. Spampinato, F. Franchetti
  Design and Specification of Large-scale Simulations for GPUs using FFTX
  Supercomputing 2019 (SC'19), Poster


• F. Franchetti, D. G. Spampinato, A. Kulkarni, D. T. Popovici, T. M. Low, M. Franusich, A. Canning, P. McCorquodale, B. Van Straalen, P. Colella
  FFTX and SpectralPack: A First Look
  IEEE International Conference on High Performance Computing, Data, and Analytics (HiPC), 2018


• F. Franchetti, T. M. Low, D. T. Popovici, R. M. Veras, D. G. Spampinato, J. R. Johnson, M. Püschel, J. C. Hoe, and J. M. F. Moura
  SPIRAL: Extreme Performance Portability
  Proceedings of the IEEE special issue on From High Level Specification to High Performance Code, Vol. 206, No. 11, 2018


• SPIRAL, FFTX, and the Path to SpectralPACK


• An Approach to Specifying and Automatically Optimizing Fourier Transform Based Operations (Supporting research)


• FFTX Documentation


• 2022 ECP Annual Meeting: Recordings and Slides

For more information, visit Franz Franchetti's projects page.

Copyrights to many of the above papers are held by the publishers. The attached PDF files are preprints. It is understood that all persons copying this information will adhere to the terms and constraints invoked by each author's copyright. These works may not be reposted without the explicit permission of the copyright holder. Some links to papers above connect to IEEE Xplore with permission from IEEE, and viewers must follow all of IEEE's copyright policies.