Cronus: A platform for parallel code generation based on computational geometry methods

Andronikos, T; Ciorba, FM; Theodoropoulos, P; Kamenopoulos, D; Papakonstantinou, G

dc.contributor.author	Andronikos, T	en
dc.contributor.author	Ciorba, FM	en
dc.contributor.author	Theodoropoulos, P	en
dc.contributor.author	Kamenopoulos, D	en
dc.contributor.author	Papakonstantinou, G	en
dc.date.accessioned	2014-03-01T01:28:05Z
dc.date.available	2014-03-01T01:28:05Z
dc.date.issued	2008	en
dc.identifier.issn	0164-1212	en
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/18706
dc.subject	Code generation	en
dc.subject	Dynamic scheduling	en
dc.subject	General loops	en
dc.subject	Shared and distributed memory architectures	en
dc.subject.classification	Computer Science, Software Engineering	en
dc.subject.classification	Computer Science, Theory & Methods	en
dc.subject.other	Chlorine compounds	en
dc.subject.other	Codes (standards)	en
dc.subject.other	Codes (symbols)	en
dc.subject.other	Computational geometry	en
dc.subject.other	Motion estimation	en
dc.subject.other	Scheduling	en
dc.subject.other	Testing	en
dc.subject.other	Case studies	en
dc.subject.other	Code generation	en
dc.subject.other	Communication minimization	en
dc.subject.other	Data locality	en
dc.subject.other	Dithering algorithms	en
dc.subject.other	Dynamic scheduling	en
dc.subject.other	Extensive testing	en
dc.subject.other	General loops	en
dc.subject.other	Generic systems	en
dc.subject.other	Loop-carried dependencies	en
dc.subject.other	Makespan	en
dc.subject.other	Motion estimation algorithms	en
dc.subject.other	Nested Loops	en
dc.subject.other	Optimal scheduling	en
dc.subject.other	Parallel coding	en
dc.subject.other	Parallelizing	en
dc.subject.other	Run-time libraries	en
dc.subject.other	Run-time scheduling	en
dc.subject.other	Scheduling policies	en
dc.subject.other	Shared and distributed memory architectures	en
dc.subject.other	Shared-memory systems	en
dc.subject.other	Test cases	en
dc.subject.other	Transitive closures	en
dc.subject.other	Scheduling algorithms	en
dc.title	Cronus: A platform for parallel code generation based on computational geometry methods	en
heal.type	journalArticle	en
heal.identifier.primary	10.1016/j.jss.2007.11.715	en
heal.identifier.secondary	http://dx.doi.org/10.1016/j.jss.2007.11.715	en
heal.language	English	en
heal.publicationDate	2008	en
heal.abstract	This paper describes CRONUS, a platform for parallelizing general nested loops. General nested loops contain complex loop bodies (assignments, conditionals, repetitions) and exhibit uniform loop-carried dependencies. The novelty Of CRONUS is twofold: ( 1) it determines the optimal scheduling hyperplane using the QuickHull algorithm, which is more efficient than previously used methods, and (2) it implements a simple and efficient dynamic rule (successive dynamic scheduling) for the runtime scheduling of the loop iterations along the optimal hyperplane. This scheduling policy enhances data locality and improves the makespan. CRONUS provides an efficient runtime library, specifically designed for communication minimization, that performs better than more generic systems, such as Berkeley UPC. Its performance was evaluated through extensive testing. Three representative case studies are examined: the Floyd-Steinberg dithering algorithm, the Transitive Closure algorithm, and the FSBM motion estimation algorithm. The experimental results corroborate the efficiency of the parallel code. The tests show speedup ranging from 1.18 (Out of the ideal 4) to 12.29 (Out of the ideal 16) on distributed-systems and 3.60 (out of 4) to 15.79 (out of 16) on shared-memory systems. CRONUS Outperforms UPC by 5-95% depending on the test case. (C) 2007 Elsevier Inc. All rights reserved.	en
heal.publisher	ELSEVIER SCIENCE INC	en
heal.journalName	Journal of Systems and Software	en
dc.identifier.doi	10.1016/j.jss.2007.11.715	en
dc.identifier.isi	ISI:000258800900009	en
dc.identifier.volume	81	en
dc.identifier.issue	8	en
dc.identifier.spage	1389	en
dc.identifier.epage	1405	en