The fast Fourier transform (FFT) is used in many domain applications—including molecular dynamics, spectrum estimation, fast convolution and correlation, signal modulation, and wireless multimedia applications, but current state-of-the-art FFT libraries are not scalable on large heterogeneous machines with many nodes.
The main objective of the ECP FFT project is to design and develop a Highly Efficient FFTs for Exascale (heFFTe) library that provides fast and robust multidimensional FFTs for large-scale heterogeneous systems with multi-core processors and hardware accelerators. HeFFTe collects and leverages existing FFT capabilities while building a sustainable FFT library that minimizes data movements, optimizes MPI communications, overlaps computations with communications, and autotunes performance on various architectures and large scale-platforms. The current heFFTe v2.0 release achieves very good scalability on pre-exascale systems, and a performance that is close to 90% of the roofline peak.
Find out more at http://icl.utk.edu/fft/