Hilbert curve scheduling: Difference between revisions

Revision as of 16:52, 26 February 2012

Revision as of 17:31, 27 February 2012

Line 1:

In [[parallel processing]], the '''Hilbert curve scheduling''' method turns a multidimensional [[task scheduling|task allocation]] problem into a one dimensional space filling problem using [[Hilbert curve]]s, assigning related tasks to locations with higher levels of proximity.<ref name=Hilbert>''Scheduling for Parallel Processing'' by Maciej Drozdowski 2009 ISBN 1848823096 page 166</ref> Other [space filling curves]] may also be used in various computing applications for similar purposes.<ref name=Hilbert/>

The [[Simple Linux Utility for Resource Management|SLURM job scheduler]] which is used on a number of [[supercomputers]] uses a best fit algorithm based on Hilbert curve scheduling in order to optimize locality of task assignments.<ref name=Eitan>''Job Scheduling Strategies for Parallel Processing:'' by Eitan Frachtenberg and Uwe Schwiegelshohn 2010 ISBN 3642046320 pages 138-144</ref>