DMCA
Optimization-Based Mapping Framework for Parallel Applications (2010)
| Venue: | University of the Basque Country |
| Citations: | 1 - 0 self |
BibTeX
@TECHREPORT{Pascual10optimization-basedmapping,
author = {Jose A. Pascual and Jose Miguel-alonso and Jose A. Lozano},
title = {Optimization-Based Mapping Framework for Parallel Applications},
institution = {University of the Basque Country},
year = {2010}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
The optimal mapping of tasks of a parallel program onto nodes of a parallel computing system has a remarkable impact on application performance. In this paper we propose an optimization framework to solve the mapping problem, which takes into account the communication matrix of the application and a cost matrix that depends on the topology of the parallel system. This cost function can be a distance matrix (the classic approach), or can take into account other considerations. We propose a novel definition of the cost criterion, applicable to torus networks, that tries to distribute traffic evenly over the different axes: the Traffic Distribution criterion. As the mapping problem is a particular instance of the Quadratic Assignment Problem, we can apply any QAP solver to the mapping. In particular, in this work we use GRASP. Using simulation, we test the performance of the optimization-based mappings, and compare it with that of trivial mappings (consecutive, random), in two different environments: static (a single applications uses all system resources all the time) and dynamic (applications are managed by a scheduler, and use and free resources as needed), always using a system with a 2D topology and real application traffic. In both environments optimization-based mappings with the TD criterion provides excellent performance. 1 1
Keyphrases
optimization-based mapping framework parallel application mapping problem qap solver distance matrix cost criterion novel definition remarkable impact cost matrix application performance optimization framework quadratic assignment problem trivial mapping system resource real application traffic classic approach optimization-based mapping account consideration parallel program parallel system td criterion free resource optimal mapping parallel computing system particular instance communication matrix traffic distribution criterion single application environment optimization-based mapping different environment different ax cost function excellent performance
