Model-Driven Choice of Numerical Methods for the Solution of the Linear Advection Equation

Abstract Designing a partial differential equations solver is a complex task which involves making choices about the solution algorithm and its parameters. Such choices are usually done on the basis of personal preference or numerical experiments, which can introduce significant bias on the selection process. In this work we develop a methodology to drive this selection process towards the optimal choices by modelling the accuracy and the performance of the solution algorithm. We show how this methodology can be successfully applied on the linear advection problem. As a result, the selection can be optimally performed with a much lower investment on the development of high-performance versions of the solvers and without using the target architecture for numerical experiments.

[1]  P. Lauritzen Numerical techniques for global atmospheric models , 2011 .

[2]  Michael Baldauf Stability analysis for linear discretisations of the advection equation with Runge-Kutta time integration , 2008, J. Comput. Phys..

[3]  David P. Lockard,et al.  High accuracy algorithms for computational aeroacoustics , 1994 .

[4]  David Joyner,et al.  Open source computer algebra systems: SymPy , 2012, ACCA.

[5]  Sally A. McKee,et al.  Hitting the memory wall: implications of the obvious , 1995, CARN.

[6]  John Thuburn,et al.  Vertical Discretizations: Some Basic Ideas , 2011 .

[7]  T. Colonius,et al.  Computational aeroacoustics: progress on nonlinear problems of sound generation , 2004 .

[8]  Matteo Bernardini,et al.  A general strategy for the optimization of Runge-Kutta schemes for wave propagation phenomena , 2009, J. Comput. Phys..

[9]  John Thuburn Horizontal Discretizations: Some Basic Ideas , 2011 .

[10]  Wim Desmet,et al.  Optimal Runge-Kutta schemes for discontinuous Galerkin space discretizations applied to wave propagation problems , 2012, J. Comput. Phys..

[11]  Steven E. Benzley,et al.  A Comparison of All Hexagonal and All Tetrahedral Finite Element Meshes for Elastic and Elasto-plastic Analysis , 2011 .

[12]  Michael Hofmann,et al.  Comparison of scalable fast methods for long-range interactions. , 2013, Physical review. E, Statistical, nonlinear, and soft matter physics.

[13]  Joel H. Ferziger,et al.  Computational methods for fluid dynamics , 1996 .

[14]  Manuel Calvo,et al.  Short note: a new minimum storage Runge-Kutta scheme for computational acoustics , 2004 .

[15]  Sergio Pirozzoli,et al.  Performance analysis and optimization of finite-difference schemes for wave propagation problems , 2007, J. Comput. Phys..

[16]  Ilaria Perugia,et al.  An A Priori Error Analysis of the Local Discontinuous Galerkin Method for Elliptic Problems , 2000, SIAM J. Numer. Anal..

[17]  C. J. Cotter,et al.  Compatible finite element methods for numerical weather prediction , 2014 .

[18]  Arturo Cifuentes,et al.  A performance study of tetrahedral and hexahedral elements in 3-D finite element structural analysis , 1992 .

[19]  D. Durran Numerical Methods for Fluid Dynamics: With Applications to Geophysics , 2010 .

[20]  David W. Zingg,et al.  Comparison of High-Accuracy Finite-Difference Methods for Linear Wave Propagation , 2000, SIAM J. Sci. Comput..

[21]  R. D. Richtmyer,et al.  Survey of the stability of linear finite difference equations , 1956 .

[22]  Samuel Williams,et al.  Roofline: an insightful visual performance model for multicore architectures , 2009, CACM.

[23]  J. Butcher The numerical analysis of ordinary differential equations: Runge-Kutta and general linear methods , 1987 .

[24]  Runge-Kutta methods with minimal dispersion and dissipation for problems arising from computational acoustics , 2003 .

[25]  W. H. Reed,et al.  Triangular mesh methods for the neutron transport equation , 1973 .

[26]  M. Nallasamy,et al.  High-Accuracy Large-Step Explicit Runge-Kutta (HALE-RK) Schemes for Computational Aeroacoustics , 2006 .