A parallel-in-time approach for accelerating direct-adjoint studies

DOI10.1016/J.JCP.2020.110033MaRDI QIDQ6337872arXivWikidataFDO

Authors C. S. Skene, M. F. Eggl, Peter J. Schmid

Publication date 1 April 2020

Published in arXiv

Full work available at URL https://doi.org/10.1016/J.JCP.2020.110033

Mathematics Subject Classification ID

Parabolic equations and parabolic systems (35Kxx) Numerical methods for partial differential equations, initial value and time-dependent initial-boundary value problems (65Mxx) Computer aspects of numerical algorithms (65Yxx)

arXiv classification

math.OCphysics.comp-ph

Abstract: Parallel-in-time methods are developed to accelerate the direct-adjoint looping procedure. Particularly, we utilize the Paraexp algorithm, previously developed to integrate equations forward in time, to accelerate the direct-adjoint looping that arises from gradient-based optimization. We consider both linear and non-linear governing equations and exploit the linear, time-varying nature of the adjoint equations. Gains in efficiency are seen across all cases, showing that a Paraexp based parallel-in-time approach is feasible for the acceleration of direct-adjoint studies. This signifies a possible approach to further increase the run-time performance for optimization studies that either cannot be parallelized in space or are at their limit of efficiency gains for a parallel-in-space approach.

This page was built for publication: A parallel-in-time approach for accelerating direct-adjoint studies

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6337872)