Show simple item record

dc.contributor.advisorChen, Zizhong
dc.contributor.authorKarlsson, Christer
dc.date.accessioned2007-01-03T08:21:16Z
dc.date.accessioned2022-02-03T11:54:23Z
dc.date.available2007-01-03T08:21:16Z
dc.date.available2022-02-03T11:54:23Z
dc.date.issued2012
dc.date.submitted2012
dc.identifierT 7093
dc.identifier.urihttps://hdl.handle.net/11124/70678
dc.description2012 Fall.
dc.descriptionIncludes illustrations (some color).
dc.descriptionIncludes bibliographical references.
dc.description.abstractIn today's high performance computing, many Message Passing Interface (MPI) programs (e.g., ScaLAPACK applications, High Performance Linpack Benchmark (HPL), and most PDE solvers based on domain decomposition methods) organize their computational processes as multidimensional Cartesian grids. Applications often need to communicate in every dimension of the Cartesian grid. While extensive optimizations have been performed on single dimensional communications such as the standard MPI collective communications, little work has been done to optimize multidimensional communications. We study the impact of the MPI process-to-core mapping on the performance of multidimensional MPI communications on Cartesian grid. While the default process-to-core mappings in today's state-of-the-art MPI implementations are often optimal for single dimensional communications, we show that they are often sub-optimal for multidimensional communications. We propose an application-level multicore-aware process-to-core re-mapping scheme that is capable of achieving optimal performance for multidimensional communication operations. The application-level solution does not require any changes to the MPI's implementations; the optimization will occur in the application layer. Experiments demonstrate that a multicore-aware process-to-core re-mapping scheme improves the performance of multidimensional MPI communications by up to 80% over the default mapping scheme on the world's current third fastest supercomputer, Jaguar, located at the Oak Ridge National Laboratory.
dc.format.mediumborn digital
dc.format.mediumdoctoral dissertations
dc.languageEnglish
dc.language.isoeng
dc.publisherColorado School of Mines. Arthur Lakes Library
dc.relation.ispartof2012 - Mines Theses & Dissertations
dc.rightsCopyright of the original work is retained by the author.
dc.subjectprocess-to-core mapping
dc.subjectmulticore
dc.subjectCartesian topology
dc.subjectcluster
dc.subjectcollective communication
dc.subjectMessage Passing Interface (MPI)
dc.titleOptimizing multi-dimensional MPI communications on multi-core architectures
dc.typeText
dc.contributor.committeememberHan, Qi
dc.contributor.committeememberMehta, Dinesh P.
dc.contributor.committeememberMunoz, David (David R.)
dc.contributor.committeememberSkokan, C. K.
thesis.degree.nameDoctor of Philosophy (Ph.D.)
thesis.degree.levelDoctoral
thesis.degree.disciplineElectrical Engineering and Computer Science
thesis.degree.grantorColorado School of Mines


Files in this item

Thumbnail
Name:
Karlsson_mines_0052E_10005.pdf
Size:
2.114Mb
Format:
PDF
Description:
Optimizing multi-dimensional MPI ...

This item appears in the following Collection(s)

Show simple item record