Show simple item record

dc.contributor.authorChavarria, Daniel
dc.date.accessioned 2017-08-02T22:02:51Z
dc.date.available 2017-08-02T22:02:51Z
dc.date.issued 2001-03-23
dc.identifier.urihttps://hdl.handle.net/1911/96286
dc.description.abstract Strategies for partitioning an application's data determine both the range of suitable parallelizations and their potential efficiency. For multi-directional line-sweep computations, multipartitioned data distributions offer better parallel efficiency and scalability than block unipartitionings. This paper describes extensions to the Rice dHPF compiler for High PerformanceFortran that enable it to support multipartitioned data distributions, and optimizations that enable dHPF to generate efficient multipartitioned code. We describe experiments applying these techniques to parallelize serial versions of the NAS SP and BT application benchmarks and show that the performance of the code generated by dHPF is within a few percent of that of hand-coded parallel implementations using multipartitioning.
dc.format.extent 11 pp
dc.language.iso eng
dc.rights You are granted permission for the noncommercial reproduction, distribution, display, and performance of this technical report in any format, but this permission is only for a period of forty-five (45) days from the most recent time that you verified that this technical report is still available from the Computer Science Department of Rice University under terms that include this permission. All other rights are reserved by the author(s).
dc.title Data-Parallel Compiler Support for Multipartitioning
dc.type Technical report
dc.date.note March 23, 2001
dc.identifier.digital TR01-374
dc.type.dcmi Text
dc.identifier.citation Chavarria, Daniel. "Data-Parallel Compiler Support for Multipartitioning." (2001) https://hdl.handle.net/1911/96286.


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record