Predicting remote reuse distance patterns in UPC applications

Steven Vormwald, W. Wang, S. Carr, S. Seidel, Z. Wang

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Scopus citations

Abstract

Current work in high productivity parallel computing has focused attention on the class of partitioned global address space (PGAS) parallel programming languages because they promise to reduce the effort required to develop parallel application codes. An important aspect in achieving good performance in PGAS languages is effective handling of remote memory references. We extend a single-threaded reuse distance model to predict memory behavior for multi-threaded UPC applications. Our model handles changes in per-thread data size as well as changes in thread mapping due to problem size increases. Our results indicate the model provides good predictions of remote memory behavior by accurately predicting changes in remote memory reuse distance as a function of the problem size and the number of threads.

Original languageEnglish (US)
Title of host publicationProceedings of the 4th Conference on Partitioned Global Address Space (PGAS) Programming Models, PGAS'10
DOIs
StatePublished - Dec 1 2010
Event4th Conference on Partitioned Global Address Space (PGAS) Programming Models, PGAS'10 - New York, NY, United States
Duration: Oct 12 2010Oct 15 2010

Publication series

NameACM International Conference Proceeding Series

Other

Other4th Conference on Partitioned Global Address Space (PGAS) Programming Models, PGAS'10
CountryUnited States
CityNew York, NY
Period10/12/1010/15/10

Keywords

  • UPC
  • cache
  • reuse distance

Fingerprint

Dive into the research topics of 'Predicting remote reuse distance patterns in UPC applications'. Together they form a unique fingerprint.

Cite this