Distributed Partial Differential Equation Solving with Julia Fast Fourier Transform Library

Author

Christopher E. Mottola, University of New Hampshire, DurhamFollow

Date of Award

Fall 2022

Project Type

Senior Honors Thesis

College or School

CEPS

Department

Computer Science

Program or Major

Computer Science

Degree Name

Bachelor of Science

First Advisor

John Gibson

Abstract

Scientific computing relies on advanced computational and mathematical techniques to solve complex problems in scientific domains. For the numerical rendering of spectral, nonlinear, and dynamic phenomena, there is a growing need for greater availability of a broad class of Fourier-based algorithms to perform large scale operations on multidimensional data in distributed and optimized ways. To this effect, the Julia programming language is new and has significant advantages compared to other common languages used in scientific computing. The research presented here formulates a basis for further development in high-performance scientific computing of periodic partial differential equations through the application of distributed Fast Fourier Transforms in Julia with the PencilFFTs.jl library.

Recommended Citation

Mottola, Christopher E., "Distributed Partial Differential Equation Solving with Julia Fast Fourier Transform Library" (2022). Honors Theses and Capstones. 694.
https://scholars.unh.edu/honors/694