Although the current practice in parallel job scheduling requires jobs to specify a particular number of requested processors, most parallel jobs are moldable, i.e. the required number of processors is flexible. This paper addresses the issue of effective selection of processor partition size for moldable jobs. The proposed scheduling strategy is shown to provide significant benefits over a rigid scheduling model and is also considerably better than a previously proposed approach to moldable job scheduling.
CITATION STYLE
Srinivasan, S., Subramani, V., Kettimuthu, R., Holenarsipur, P., & Sadayappan, P. (2002). Effective selection of partition sizes for moldable scheduling of parallel jobs. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 2552, pp. 174–183). Springer Verlag. https://doi.org/10.1007/3-540-36265-7_17
Mendeley helps you to discover research relevant for your work.