Optimal Parameter Values for a Parallel Structured Adaptive Mesh Refinement Algorithm

Authors:
Michael Thuné;Stefan Söderberg
Affiliations:
-;-
Venue:
PARA '00 Proceedings of the 5th International Workshop on Applied Parallel Computing, New Paradigms for HPC in Industry and Academia
Year:
2000

Citing 6
Cited 0

A solution-adaptive upwind scheme for ideal magnetohydrodynamics

Journal of Computational Physics
Partitioning strategies for structured multiblock grids

Parallel Computing - Special issue on graph partioning and parallel computing
A common data management infrastructure for adaptive algorithms for PDE solutions

SC '97 Proceedings of the 1997 ACM/IEEE conference on Supercomputing
Numerical Mathematics and Computing

Numerical Mathematics and Computing
Comparison of Partitioning Strategies for PDE Solvers on Multiblock Grids

PARA '98 Proceedings of the 4th International Workshop on Applied Parallel Computing, Large Scale Scientific and Industrial Problems
Local adaptive mesh refinement for shock hydrodynamics

Journal of Computational Physics

Quantified Score

Hi-index	0.00

Visualization

Abstract

A blockwise approach to parallel structured adaptive mesh refinement is considered. The initial, coarse grid is divided into n1 × n2 blocks. Subsequent refinements are carried out with respect to entire blocks. The paper addresses the issue of choosing n1 and n2 optimally. A theoretical model for the execution time is formulated. Subsequently, it is suggested how to minimize the execution time with respect to the number of blocks. The approach is validated for test cases, where it successfully predicts the optimal choice of granularity. Finally, it is discussed how this can be automatized and integrated into the SAMR code.