Welcome to the ADMT Publication Server

Repartitioning Strategies for Massively Parallel Simulation of Reacting Flow

DocUID: 2015-011 Full Text: PDF

Author: Patrick H. Pisciuneri, Angen Zheng, Peyman Givi, Alexandros Labrinidis, Panos K. Chrysanthis

Abstract: The majority of parallel CFD simulators partition the domain into equal regions and assign the calculations for a particular region to a unique processor. This type of domain decomposition is vital to the efficiency of the solver. However, as the simulation develops, the workload among the partitions often become uneven (e.g., by adaptive mesh refinement, or chemically reacting regions) and a new partition should be considered. The process of repartitioning adjusts the current partition to evenly distribute the load again. We compare two repartitioning tools: Zoltan, an architecture-agnostic graph repartitioner developed at the Sandia National Laboratories; and Paragon, an architecture-aware graph repartitioner developed at the University of Pittsburgh. The comparative assessment is conducted via simulation of the Taylor-Green vortex flow with chemical reaction

Published In: 68th Annual Meeting of the APS Division of Fluid Dynamics

Volume: 60Number: 21Pages: 503

Year Published: 2015

Project: Paragon,  AragonLB Subject Area: Scientific Data Management

Publication Type: Extended Abstract

Sponsor: NSF CBET-1250171, NSF OIA-1028162

Citation:Text Latex BibTex XML Patrick H. Pisciuneri, Angen Zheng, Peyman Givi, Alexandros Labrinidis, and Panos K. Chrysanthis. Repartitioning Strategies for Massively Parallel Simulation of Reacting Flow. 68th Annual Meeting of the APS Division of Fluid Dynamics. 60(21):503. 2015.