Abstract
In this paper three strategies are described to restore dynamically the load balancing in parallel active set optimization algorithms. The efficiency of our proposals is shown by comparison with other heuristics described in related works, such as the classical Bestfit and Worstfit methods. The computational cost due to the load unbalancing in the parallel code and the communication overheads associated with the most efficient load balancing strategy are analyzed and compared in order to establish whether the distribution is convenient or not. Experimental results on a distributed memory system, the Fujitsu AP3000, highlight the accuracy of our estimations.
This work was supported by CICYT under grants TIC 2002/750 and TIC 2001-3694-C02.
Chapter PDF
Similar content being viewed by others
Keywords
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
References
Blanco, V., et al.: Performance of Parallel Iterative Solvers: a Library, a Prediction Model and a Visualization Tool. Journal of Information Science and Engineering 8(5), 763–785 (2002)
Gill, P.E., Murray, W., Wright, M.H.: Practical Optimization. Academic Press, London (1981)
Gill, P.E., Murray, W., Wright, M.H.: Numerical Linear Algebra and Optimization, vol. 1. Addison-Wesley Publishing Company, California (1991)
Haglin, D.J., Ford, R.W.: The Message-Minimizing Load Redistribution Problem. Journal of Universal Computer Science 7(4), 291–306 (2001)
Martello, S., Toth, P.: Knapsack Problems: Algorithms and Computer Implementations. Wiley Interscience Series in Discrete Mathematics and Optimization. John Wiley & Sons, New york (1990)
Pardines, I., Rivera, F.F.: Parallel Quasi-Newton Optimization on Distributed Memory Multiprocessors. Parallel Computing: Advances and Current Issues. In: Proceedings of ParCo 2001, pp. 338–345. Imperial College Press, London (2002)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2003 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Pardines, I., Rivera, F.F. (2003). Efficient Dynamic Load Balancing Strategies for Parallel Active Set Optimization Methods. In: Kosch, H., Böszörményi, L., Hellwagner, H. (eds) Euro-Par 2003 Parallel Processing. Euro-Par 2003. Lecture Notes in Computer Science, vol 2790. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-45209-6_31
Download citation
DOI: https://doi.org/10.1007/978-3-540-45209-6_31
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-40788-1
Online ISBN: 978-3-540-45209-6
eBook Packages: Springer Book Archive