File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Article: Decentralized remapping of data parallel applications in distributed memory multiprocessors

TitleDecentralized remapping of data parallel applications in distributed memory multiprocessors
Authors
Issue Date1997
PublisherJohn Wiley & Sons Ltd. The Journal's web site is located at http://www.interscience.wiley.com/jpages/1532-0626/
Citation
Concurrency Practice And Experience, 1997, v. 9 n. 12, p. 1351-1376 How to Cite?
AbstractIn this paper we present a decentralized remapping method for data parallel applications on distributed memory multiprocessors. The method uses a generalized dimension exchange (GDE) algorithm periodically during the execution of an application to balance (remap) the system's workload. We implemented this remapping method in parallel WaTor simulations and parallel image thinning applications, and found it to be effective in reducing the computation time. The average performance gain is about 20% in the WaTor simulation of a 256 ×256 ocean grid on 16 processors, and up to 8% in the thinning of a typical image of size 128 × 128 on eight processors. The performance gains due to remapping in the image thinning case are reasonably substantial given the fact that the application by its very nature does not necessarily favor remapping. We also implemented this remapping method, using up to 32 processors, for partitioning and re-partitioning of grids in computational fluid dynamics. It was found that the GDE-based parallel refinement policy, coupled with simple geometric strategies, produces partitions that are comparable in quality to those from the best serial algorithms. ©1997 John Wiley & Sons, Ltd.
Persistent Identifierhttp://hdl.handle.net/10722/89097
ISSN
2002 Impact Factor: 0.56
References

 

DC FieldValueLanguage
dc.contributor.authorXu, Cen_HK
dc.contributor.authorLau, FCMen_HK
dc.contributor.authorDiekmann, Ren_HK
dc.date.accessioned2010-09-06T09:52:21Z-
dc.date.available2010-09-06T09:52:21Z-
dc.date.issued1997en_HK
dc.identifier.citationConcurrency Practice And Experience, 1997, v. 9 n. 12, p. 1351-1376en_HK
dc.identifier.issn1040-3108en_HK
dc.identifier.urihttp://hdl.handle.net/10722/89097-
dc.description.abstractIn this paper we present a decentralized remapping method for data parallel applications on distributed memory multiprocessors. The method uses a generalized dimension exchange (GDE) algorithm periodically during the execution of an application to balance (remap) the system's workload. We implemented this remapping method in parallel WaTor simulations and parallel image thinning applications, and found it to be effective in reducing the computation time. The average performance gain is about 20% in the WaTor simulation of a 256 ×256 ocean grid on 16 processors, and up to 8% in the thinning of a typical image of size 128 × 128 on eight processors. The performance gains due to remapping in the image thinning case are reasonably substantial given the fact that the application by its very nature does not necessarily favor remapping. We also implemented this remapping method, using up to 32 processors, for partitioning and re-partitioning of grids in computational fluid dynamics. It was found that the GDE-based parallel refinement policy, coupled with simple geometric strategies, produces partitions that are comparable in quality to those from the best serial algorithms. ©1997 John Wiley & Sons, Ltd.en_HK
dc.languageengen_HK
dc.publisherJohn Wiley & Sons Ltd. The Journal's web site is located at http://www.interscience.wiley.com/jpages/1532-0626/en_HK
dc.relation.ispartofConcurrency Practice and Experienceen_HK
dc.rightsConcurrency and Computation: Practice & Experience. Copyright © John Wiley & Sons Ltd.en_HK
dc.titleDecentralized remapping of data parallel applications in distributed memory multiprocessorsen_HK
dc.typeArticleen_HK
dc.identifier.openurlhttp://library.hku.hk:4550/resserv?sid=HKU:IR&issn=1532-0626&volume=9&spage=1351&epage=1376&date=1997&atitle=Decentralized+Remapping+of+Data+Parallel+Applications+in+Distributed+Memory+Multiprocessorsen_HK
dc.identifier.emailLau, FCM:fcmlau@cs.hku.hken_HK
dc.identifier.authorityLau, FCM=rp00221en_HK
dc.description.naturelink_to_subscribed_fulltext-
dc.identifier.scopuseid_2-s2.0-0031383079en_HK
dc.identifier.hkuros30044en_HK
dc.relation.referenceshttp://www.scopus.com/mlt/select.url?eid=2-s2.0-0031383079&selection=ref&src=s&origin=recordpageen_HK
dc.identifier.volume9en_HK
dc.identifier.issue12en_HK
dc.identifier.spage1351en_HK
dc.identifier.epage1376en_HK
dc.identifier.scopusauthoridXu, C=8701888000en_HK
dc.identifier.scopusauthoridLau, FCM=7102749723en_HK
dc.identifier.scopusauthoridDiekmann, R=6603859330en_HK

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats