XMDS2: Fast, scalable simulation of coupled stochastic partial differential equations

Abstract

XMDS2 is a cross-platform, GPL-licensed, open source package for numerically integrating initial value problems that range from a single ordinary differential equation up to systems of coupled stochastic partial differential equations. The equations are described in a high-level XML-based script, and the package generates low-level optionally parallelised C++ code for the efficient solution of those equations. It combines the advantages of high-level simulations, namely fast and low-error development, with the speed, portability and scalability of hand-written code. XMDS2 is a complete redesign of the XMDS package, and features support for a much wider problem space while also producing faster code. Program summaryProgram title: XMDS2Catalogue identifier: AENK_v1_0Program summary URL:http://cpc.cs.qub.ac.uk/summaries/AENK_v1_0.htmlProgram obtainable from: CPC Program Library, Queen’s University, Belfast, N. IrelandLicensing provisions: GNU General Public License, version 2No. of lines in distributed program, including test data, etc.: 872490No. of bytes in distributed program, including test data, etc.: 45522370Distribution format: tar.gzProgramming language: Python and C++.Computer: Any computer with a Unix-like system, a C++ compiler and Python.Operating system: Any Unix-like system; developed under Mac OS X and GNU/Linux.RAM: Problem dependent (roughly 50 bytes per grid point)Classification: 4.3, 6.5.External routines: The external libraries required are problem-dependent. Uses FFTW3 Fourier transforms (used only for FFT-based spectral methods), dSFMT random number generation (used only for stochastic problems), MPI message-passing interface (used only for distributed problems), HDF5, GNU Scientific Library (used only for Bessel-based spectral methods) and a BLAS implementation (used only for non-FFT-based spectral methods).Nature of problem: General coupled initial-value stochastic partial differential equations.Solution method: Spectral method with method-of-lines integrationRunning time: Determined by the size of the problem