SpiralGen maintains several open source packages on https://github.com/spiral-software.
SPIRAL is a performance engineering and algorithm optimization toolkit that allows users to take full advantage of the performance capabilities of hardware. SPIRAL delivers efficient implementations of computationally intensive mathematics that enable near achievable peak hardware performance with minimal programming effort.
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 discrete Fourier transforms. Though backward compatible with FFTW, this is an entirely new work developed as a cooperative effort between Lawrence Berkeley National Laboratory, Carnegie Mellon University, and SpiralGen, under the Department of Energy’s Exascale Computing Project.Learn More
FFTX for Python
The package provides a single API for several FFTs, along with some convolution-like transforms, that run on either CPUs or GPUs (NVIDIA and AMD) and supports single and double precision and both C (row-major) and Fortran (column-major) arrays. The API uses the dimensions, datatype, ordering, and location of an array to determine which variant of a transform to invoke, simplifying application code intended for multiple target environments.Learn More