Scalable work stealing

  • Authors:

  • James Dinan;D. Brian Larkins;P. Sadayappan;Sriram Krishnamoorthy;Jarek Nieplocha

  • Affiliations:

  • The Ohio State University, Columbus, OH;The Ohio State University, Columbus, OH;The Ohio State University, Columbus, OH;Pacific Northwest National Laboratory, Richland, WA;Pacific Northwest National Laboratory, Richland, WA

  • Venue:
  • Proceedings of the Conference on High Performance Computing Networking, Storage and Analysis
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

Irregular and dynamic parallel applications pose significant challenges to achieving scalable performance on large-scale multicore clusters. These applications often require ongoing, dynamic load balancing in order to maintain efficiency. Scalable dynamic load balancing on large clusters is a challenging problem which can be addressed with distributed dynamic load balancing systems. Work stealing is a popular approach to distributed dynamic load balancing; however its performance on large-scale clusters is not well understood. Prior work on work stealing has largely focused on shared memory machines. In this work we investigate the design and scalability of work stealing on modern distributed memory systems. We demonstrate high efficiency and low overhead when scaling to 8,192 processors for three benchmark codes: a producer-consumer benchmark, the unbalanced tree search benchmark, and a multiresolution analysis kernel.