Efficient Algorithms for Scheduling Moldable Tasks
Abstract
We study the problem of scheduling independent moldable tasks on processors that arises in large-scale parallel computations. When tasks are monotonic, the best known result is a -approximation algorithm for makespan minimization with a complexity linear in and polynomial in and where is arbitrarily small. We propose a new perspective of the existing speedup models: the speedup of a task is linear when the number of assigned processors is small (up to a threshold ) while it presents monotonicity when ranges in ; the bound indicates an unacceptable overhead when parallelizing on too many processors. The generality of this model is proved to be between the classic monotonic and linear-speedup models. For any given integer , let . In this paper, we propose a -approximation algorithm for makespan minimization where (). As a by-product, we also propose a -approximation algorithm for throughput maximization with a common deadline.
Cite
@article{arxiv.1609.08588,
title = {Efficient Algorithms for Scheduling Moldable Tasks},
author = {Xiaohu Wu and Patrick Loiseau},
journal= {arXiv preprint arXiv:1609.08588},
year = {2023}
}
Comments
Accepted by European Journal of Operational Research