Interactive Locality Optimization on NUMA Architectures

Tao Mu, Jie Tao, Martin Schulz, Sally A. McKee

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

13 Scopus citations

Abstract

Optimizing the performance of shared-memory NUMA programs remains something of a black art, requiring that application writers possess deep understanding of their programs' behaviors. This difficulty represents one of the remaining hindrances to the widespread adoption and deployment of these cost-efficient and scalable shared-memory NUMA architectures. To address this problem, we have developed a performance monitoring infrastructure and a corresponding set of tools to aid in visualizing and understanding the subtleties of the memory access behavior of parallel NUMA applications with large datasets. The tools are designed to be general, interoperable, and easily portable. We give detailed examples of the use of one particular tool in the set. We have used this memory access visualization tool profitably on a range of applications, improving performance by around 90%, on average.

Original languageEnglish
Title of host publicationProceedings of ACM Symposium on Software Visualization
EditorsS.N. Spencer
PublisherAssociation for Computing Machinery (ACM)
Pages133-141
Number of pages9
ISBN (Print)1581136420, 9781581136425
DOIs
StatePublished - 2003
EventProceedings of the ACM 2003 Symposium on Software Visualization (SoftVis 2003) - San Diego, CA, United States
Duration: 11 Jun 200313 Jun 2003

Publication series

NameProceedings of ACM Symposium on Software Visualization

Conference

ConferenceProceedings of the ACM 2003 Symposium on Software Visualization (SoftVis 2003)
Country/TerritoryUnited States
CitySan Diego, CA
Period11/06/0313/06/03

Keywords

  • Distributed Systems
  • Interactive Locality Optimizations
  • NUMA Architectures
  • Performance Visualization

Fingerprint

Dive into the research topics of 'Interactive Locality Optimization on NUMA Architectures'. Together they form a unique fingerprint.

Cite this