DMCA
Parallel Cluster Identification for Multidimensional Lattices (1995)
| Venue: | IEEE Transactions on Parallel and Distributed Systems |
| Citations: | 4 - 2 self |
BibTeX
@ARTICLE{Fink95parallelcluster,
author = {Stephen J. Fink and Craig Huston and Scott B. Baden and Karl Jansen},
title = {Parallel Cluster Identification for Multidimensional Lattices},
journal = {IEEE Transactions on Parallel and Distributed Systems},
year = {1995},
volume = {8}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
The cluster identification problem is a variant of connected component labeling that arises in cluster algorithms for spin models in statistical physics. We present a multidimensional version of Belkhale and Banerjee's Quad algorithm for connected component labeling on distributed memory parallel computers. Our extension abstracts away extraneous spatial connectivity information in more than two dimensions, simplifying implementation for higher dimensionality. We identify two types of locality present in cluster configurations, and present optimizations to exploit locality for better performance. Performance results from 2D, 3D, and 4D Ising model simulations with Swendson-Wang dynamics show that the optimizations improve performance by 20-80%. Index terms: cluster identification, Ising model, connected component labeling, parallel algorithm, Swendson-Wang dynamics 1 1 Introduction The cluster identification problem is a variant of connected component labeling that arises in cluste...
Keyphrases
connected component parallel cluster identification multidimensional lattice swendson-wang dynamic cluster identification problem quad algorithm cluster algorithm present optimization multidimensional version component labeling cluster identification performance result index term statistical physic extraneous spatial connectivity information locality present distributed memory parallel computer ising model cluster configuration parallel algorithm spin model ising model simulation
