Show simple item record

A Lossy, Synchronization-Free, Race-Full, But Still Acceptably Accurate Parallel Space-Subdivision Tree Construction Algorithm

dc.date.accessioned2012-02-23T20:30:03Z
dc.date.accessioned2018-11-26T22:26:47Z
dc.date.available2012-02-23T20:30:03Z
dc.date.available2018-11-26T22:26:47Z
dc.date.issued2012-02-23
dc.identifier.urihttp://hdl.handle.net/1721.1/69177
dc.identifier.urihttp://repository.aust.edu.ng/xmlui/handle/1721.1/69177
dc.description.abstractWe present a new synchronization-free space-subdivision tree construction algorithm. Despite data races, this algorithm produces trees that are consistent enough for the client Barnes-Hut center of mass and force computation phases to use successfully. Our performance results show that eliminating synchronization improves the performance of the parallel algorithm by approximately 20%. End-to-end accuracy results show that the resulting partial data structure corruption has a neglible effect on the overall accuracy of the Barnes-Hut N-body simulation. We note that many data structure manipulation algorithms use many of the same basic operations (linked data structure updates and array insertions) as our tree construction algorithm. We therefore anticipate that the basic principles the we develop in this paper may effectively guide future efforts in this area.en_US
dc.format.extent12 p.en_US
dc.subjectData Structure Repairen_US
dc.subjectData Racesen_US
dc.subjectAcceptabilityen_US
dc.subjectParallel Computingen_US
dc.subjectSynchronizationen_US
dc.titleA Lossy, Synchronization-Free, Race-Full, But Still Acceptably Accurate Parallel Space-Subdivision Tree Construction Algorithmen_US


Files in this item

FilesSizeFormatView
MIT-CSAIL-TR-2012-005.pdf220.6Kbapplication/pdfView/Open

This item appears in the following Collection(s)

Show simple item record