Cartesian partitioning models for 2D and 3D parallel SpGEMM algorithms

buir.contributor.authorDemirci, Gündüz Vehbi
buir.contributor.authorAykanat, Cevdet
dc.citation.epage2775en_US
dc.citation.issueNumber12en_US
dc.citation.spage2763en_US
dc.citation.volumeNumber31en_US
dc.contributor.authorDemirci, Gündüz Vehbien_US
dc.contributor.authorAykanat, Cevdeten_US
dc.date.accessioned2021-02-11T11:32:26Z
dc.date.available2021-02-11T11:32:26Z
dc.date.issued2020
dc.departmentDepartment of Computer Engineeringen_US
dc.description.abstractThe focus is distributed-memory parallelization of sparse-general-matrix-multiplication (SpGEMM). Parallel SpGEMM algorithms are classified under one-dimensional (1D), 2D, and 3D categories denoting the number of dimensions by which the 3D sparse workcube representing the iteration space of SpGEMM is partitioned. Recently proposed successful 2D- and 3D-parallel SpGEMM algorithms benefit from upper bounds on communication overheads enforced by 2D and 3D cartesian partitioning of the workcube on 2D and 3D virtual processor grids, respectively. However, these methods are based on random cartesian partitioning and do not utilize sparsity patterns of SpGEMM instances for reducing the communication overheads. We propose hypergraph models for 2D and 3D cartesian partitioning of the workcube for further reducing the communication overheads of these 2D- and 3D- parallel SpGEMM algorithms. The proposed models utilize two- and three-phase partitioning that exploit multi-constraint hypergraph partitioning formulations. Extensive experimentation performed on 20 SpGEMM instances by using upto 900 processors demonstrate that proposed partitioning models significantly improve the scalability of 2D and 3D algorithms. For example, in 2D-parallel SpGEMM algorithm on 900 processors, the proposed partitioning model respectively achieves 85 and 42 percent decrease in total volume and total number of messages, leading to 1.63 times higher speedup compared to random partitioning, on average.en_US
dc.identifier.doi10.1109/TPDS.2020.3000708en_US
dc.identifier.issn1045-9219
dc.identifier.urihttp://hdl.handle.net/11693/55081
dc.language.isoEnglishen_US
dc.publisherIEEEen_US
dc.relation.isversionofhttps://dx.doi.org/10.1109/TPDS.2020.3000708en_US
dc.source.titleIEEE Transactions on Parallel and Distributed Systemsen_US
dc.subjectSparse matrix-matrix multiplicationen_US
dc.subjectSpGEMMen_US
dc.subjectSparse SUMMA SpGEMMen_US
dc.subjectSplit-3D-SpGEMMen_US
dc.subjectHypergraph partitioningen_US
dc.subjectCommunication costen_US
dc.subjectBandwidthen_US
dc.subjectLatencyen_US
dc.titleCartesian partitioning models for 2D and 3D parallel SpGEMM algorithmsen_US
dc.typeArticleen_US
Files
Original bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
Cartesian_Partitioning_Models_for_2D_and_3D_Parallel_SpGEMM_Algorithms.pdf
Size:
1.3 MB
Format:
Adobe Portable Document Format
Description:
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
1.71 KB
Format:
Item-specific license agreed upon to submission
Description: