Pubs2

MapReduce for the Cell B.E. Architecture

| Sorted by Date | Classified by Publication Type | Classified by Project |

Marc de Kruijf and Karthikeyan Sankaralingam. MapReduce for the Cell B.E. Architecture. IBM Journal of Research and Development, 53(5), 2009.

Download

(unavailable)

Abstract

MapReduce is a simple and flexible parallel programming model proposedby Google for large-scale distributed data processing. In this paper,we present a design and prototype implementation of MapReduce for theCell Broadband Engine Architecture (CBEA). The MapReduce modelprovides a simple machine abstraction that shi elds users fromparallelization and other distributed programming complications. Thegoal of this paper is t o describe the tradeoffs in the design of theruntime and demonstrate the potential for high performance. We studythe basic characteristics of the MapReduce model and identify threetypes of MapReduce applications: map dominated, partition dominated,and sort dominated. We evaluate our runtime performance, scalability,an d efficiency for microbenchmarks representing each of theseapplication types as well as for complete applic ations. We find thatmap-dominated applications map well to the CBEA and that our prototypesustains high pe rformance on these applications. Forpartition-dominated and sort-dominated applications, we analyzeruntime performance, identify sources of inefficiency, and proposeseveral future enhancements to significantly imp roveperformance. Overall, we find that the simplicity and efficiency ofthe model make it an attractive too l for programming Cell BroadbandEngine processor-based platforms.

BibTeX

 @article{ibmjr09:MapreduceCell2007,
   AUTHOR = {Marc de Kruijf and Karthikeyan Sankaralingam},
   TITLE = "{MapReduce for the Cell B.E. Architecture}",
   abstract = {
 MapReduce is a simple and flexible parallel programming model proposed
 by Google for large-scale distributed data processing. In this paper,
 we present a design and prototype implementation of MapReduce for the
 Cell Broadband Engine<AE> Architecture (CBEA). The MapReduce model
 provides a simple machine abstraction that shi elds users from
 parallelization and other distributed programming complications. The
 goal of this paper is t o describe the tradeoffs in the design of the
 runtime and demonstrate the potential for high performance. We study
 the basic characteristics of the MapReduce model and identify three
 types of MapReduce applications: map dominated, partition dominated,
 and sort dominated. We evaluate our runtime performance, scalability,
 an d efficiency for microbenchmarks representing each of these
 application types as well as for complete applic ations. We find that
 map-dominated applications map well to the CBEA and that our prototype
 sustains high pe rformance on these applications. For
 partition-dominated and sort-dominated applications, we analyze
 runtime performance, identify sources of inefficiency, and propose
 several future enhancements to significantly imp rove
 performance. Overall, we find that the simplicity and efficiency of
 the model make it an attractive too l for programming Cell Broadband
 Engine processor-based platforms. },
  journal = {IBM Journal of Research and Development},
  volume = {53},
  number = {5},
  year = {2009},
  bib_pubtype = {Journal},
  bib_rescat = {proj-opinion}
 }

Generated by bib.pl (written by Patrick Riley ) on Sun Sep 26, 2021 16:14:28 time=1207019082