Parallel Performance Optimization of Large-Scale Unstructured Data Visualization for the Earth Simulator
Abstract
This paper describes some efficient parallel performance optimization strategies for large-scale unstructured data visualization on SMP cluster machines including the Earth Simulator in Japan. The three-level hybrid parallelization is employed in our implementation, consisting of message passing for inter-SMP node communication, loop directives by OpenMP for intra-SMP node parallelization, and vectorization for each processing element (PE). In order to improve the speedup performance for the hybrid parallelization, some techniques, such as multi-coloring for removing data race and dynamic load repartition for load balancing, are considered. Good visualization images and high parallel performance have been achieved on Hitachi SR8000 for large-scale unstructured datasets, which shows the feasibility and effectiveness of our strategies.
BibTeX
@inproceedings {10.2312:EGPGV:EGPGV02:133-140,
booktitle = {Eurographics Workshop on Parallel Graphics and Visualization},
editor = {D. Bartz and X. Pueyo and E. Reinhard},
title = {{Parallel Performance Optimization of Large-Scale Unstructured Data Visualization for the Earth Simulator}},
author = {Chen, L. and Fujishiro, I. and Nakajima, K.},
year = {2002},
publisher = {The Eurographics Association},
ISSN = {1727-348X},
ISBN = {1-58113-579-3},
DOI = {10.2312/EGPGV/EGPGV02/133-140}
}
booktitle = {Eurographics Workshop on Parallel Graphics and Visualization},
editor = {D. Bartz and X. Pueyo and E. Reinhard},
title = {{Parallel Performance Optimization of Large-Scale Unstructured Data Visualization for the Earth Simulator}},
author = {Chen, L. and Fujishiro, I. and Nakajima, K.},
year = {2002},
publisher = {The Eurographics Association},
ISSN = {1727-348X},
ISBN = {1-58113-579-3},
DOI = {10.2312/EGPGV/EGPGV02/133-140}
}