Performance of preconditioners for large-scale simulations using nek5000

Offermans N, Peplinski A, Marin O, Merzari E, Schlatter P (2020)


Publication Type: Conference contribution

Publication year: 2020

Journal

Publisher: Springer

Book Volume: 134

Pages Range: 263-272

Conference Proceedings Title: Lecture Notes in Computational Science and Engineering

Event location: London, GBR

ISBN: 9783030396466

DOI: 10.1007/978-3-030-39647-3_20

Abstract

BoomerAMG, the algebraic multigrid solver from the hypre library, is used to solve a coarse grid problem which is part of the preconditioning strategy for the pressure equation arising from the numerical resolution of the Navier–Stokes equations. A set of optimal parameters for the setup phase is determined and used for selected strong scaling tests on two different supercomputers, namely Mira and Hazel Hen, on up to 131, 072 compute cores. The results are compared to an existing algebraic multigrid solver, designed specifically for the coarse grid problem at hand. It is shown that the BoomerAMG solver is fast and scalable, and that performance depends on the computer architecture. The test cases considered are the turbulent flow past a NACA4412 airfoil and the turbulent flow inside wire-tapped pin bundles.

Authors with CRIS profile

Involved external institutions

How to cite

APA:

Offermans, N., Peplinski, A., Marin, O., Merzari, E., & Schlatter, P. (2020). Performance of preconditioners for large-scale simulations using nek5000. In Spencer J. Sherwin, Joaquim Peiró, Peter E. Vincent, David Moxey, Christoph Schwab (Eds.), Lecture Notes in Computational Science and Engineering (pp. 263-272). London, GBR: Springer.

MLA:

Offermans, N., et al. "Performance of preconditioners for large-scale simulations using nek5000." Proceedings of the 12th International Conference on Spectral and High-Order Methods, ICOSAHOM 2018, London, GBR Ed. Spencer J. Sherwin, Joaquim Peiró, Peter E. Vincent, David Moxey, Christoph Schwab, Springer, 2020. 263-272.

BibTeX: Download