Loading [MathJax]/extensions/Safe.js
SciPost logo

SciPost Submission Page

cuPSS: a package for the pseudo-spectral integration of stochastic PDEs

by Fernando Caballero

Submission summary

Authors (as registered SciPost users): Fernando Caballero
Submission information
Preprint Link: scipost_202503_00045v1  (pdf)
Code repository: https://github.com/fcaballerop/cuPSS/tree/main
Code version: v0.2025.1
Code license: MIT
Date submitted: 2025-03-25 01:10
Submitted by: Caballero, Fernando
Submitted to: SciPost Physics Codebases
Ontological classification
Academic field: Physics
Specialties:
  • Condensed Matter Physics - Computational
  • Statistical and Soft Matter Physics
  • Active Matter
Approach: Computational

Abstract

This manuscript introduces cuPSS, a package to numerically integrate arbitrary systems of stochastic partial differential equations (SPDEs). SPDEs are a powerful mathematical tool to describe a large class of physical systems. Their numerical integration usually relies eiher on in-house made codes or external packages like MATLAB, Mathematica, Fenicsx, OpenFOAM, Dedalus, and others. These packages rarely offer a good combination of speed, generality, and the option to easily add stochasticity to the system, while in-house codes depend on certain expertise to obtain good performance, and are usually written for each specific use-case, sacrificing modularity and reusability. cuPSS, by contrast, can deal with the integration of an arbitrary number of stochastic fields and differential equations, and is written in CUDA C++, thus enabling by default GPU acceleration. The numerical integration is done pseudo-spectrally in flat lattices in one, two and three dimensions. This manuscript describes the basic functionality of cuPSS, with an example and benchmarking, showing that cuPSS offers a considerable improvement in speed over other popular finite-difference and spectral solvers.

Current status:
In refereeing

Login to report or comment