skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Parallel tetrahedral mesh adaptation with dynamic load balancing

Abstract

The ability to dynamically adapt an unstructured grid is a powerful tool for efficiently solving computational problems with evolving physical features. In this paper, we report on our experience parallelizing an edge-based adaptation scheme, called 3D-TAG, using message passing. Results show excellent speedup when a realistic helicopter rotor mesh is randomly refined. However, performance deteriorates when the mesh is refined using a solution-based error indicator since mesh adaptation for practical problems occurs in a localized region, creating a severe load imbalance. To address this problem, we have developed PLUM, a global dynamic load balancing framework for adaptive numerical computations. Even though PLUM primarily balances processor workloads for the solution phase, it reduces the load imbalance problem within mesh adaptation by repartitioning the mesh after targeting edges for refinement but before the actual subdivision. This dramatically improves the performance of parallel 3D-TAG since refinement occurs in a more load balanced fashion. We also present optimal and heuristic algorithms that, when applied to the default mapping of a parallel repartitioner, significantly reduce the data redistribution overhead. Finally, portability is examined by comparing performance on three state-of-the-art parallel machines.

Authors:
; ;
Publication Date:
Research Org.:
Lawrence Berkeley National Lab. (LBNL), Berkeley, CA (United States)
Sponsoring Org.:
USDOE Director, Office of Science. Office of Advanced Scientific Computing Research (US)
OSTI Identifier:
776616
Report Number(s):
LBNL-46244
R&D Project: 618110; TRN: AH200118%%454
DOE Contract Number:  
AC03-76SF00098
Resource Type:
Journal Article
Journal Name:
Parallel Computing Journal
Additional Journal Information:
Journal Volume: 26; Journal Issue: 12; Other Information: Journal Publication Date: Nov. 2000; PBD: 28 Jun 2000
Country of Publication:
United States
Language:
English
Subject:
99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE; PARALLEL PROCESSING; ALGORITHMS; HELICOPTERS; PERFORMANCE; ROTORS; DATA TRANSMISSION; MESH GENERATION; P CODES; LOAD MANAGEMENT; NUMBER CODES

Citation Formats

Oliker, Leonid, Biswas, Rupak, and Gabow, Harold N. Parallel tetrahedral mesh adaptation with dynamic load balancing. United States: N. p., 2000. Web. doi:10.1016/S0167-8191(00)00047-8.
Oliker, Leonid, Biswas, Rupak, & Gabow, Harold N. Parallel tetrahedral mesh adaptation with dynamic load balancing. United States. https://doi.org/10.1016/S0167-8191(00)00047-8
Oliker, Leonid, Biswas, Rupak, and Gabow, Harold N. 2000. "Parallel tetrahedral mesh adaptation with dynamic load balancing". United States. https://doi.org/10.1016/S0167-8191(00)00047-8.
@article{osti_776616,
title = {Parallel tetrahedral mesh adaptation with dynamic load balancing},
author = {Oliker, Leonid and Biswas, Rupak and Gabow, Harold N},
abstractNote = {The ability to dynamically adapt an unstructured grid is a powerful tool for efficiently solving computational problems with evolving physical features. In this paper, we report on our experience parallelizing an edge-based adaptation scheme, called 3D-TAG, using message passing. Results show excellent speedup when a realistic helicopter rotor mesh is randomly refined. However, performance deteriorates when the mesh is refined using a solution-based error indicator since mesh adaptation for practical problems occurs in a localized region, creating a severe load imbalance. To address this problem, we have developed PLUM, a global dynamic load balancing framework for adaptive numerical computations. Even though PLUM primarily balances processor workloads for the solution phase, it reduces the load imbalance problem within mesh adaptation by repartitioning the mesh after targeting edges for refinement but before the actual subdivision. This dramatically improves the performance of parallel 3D-TAG since refinement occurs in a more load balanced fashion. We also present optimal and heuristic algorithms that, when applied to the default mapping of a parallel repartitioner, significantly reduce the data redistribution overhead. Finally, portability is examined by comparing performance on three state-of-the-art parallel machines.},
doi = {10.1016/S0167-8191(00)00047-8},
url = {https://www.osti.gov/biblio/776616}, journal = {Parallel Computing Journal},
number = 12,
volume = 26,
place = {United States},
year = {Wed Jun 28 00:00:00 EDT 2000},
month = {Wed Jun 28 00:00:00 EDT 2000}
}