Optimizing bandwidth limited problems using one-sided communication and overlap

  • Authors:

  • Christian Bell;Dan Bonachea;Rajesh Nishtala;Katherine Yelick

  • Affiliations:

  • University of California, Berkeley, Berkeley, CA and Lawrence Berkeley National Laboratory, Berkeley, CA;University of California, Berkeley, Berkeley, CA;University of California, Berkeley, Berkeley, CA;University of California, Berkeley, Berkeley, CA and Lawrence Berkeley National Laboratory, Berkeley, CA

  • Venue:
  • IPDPS'06 Proceedings of the 20th international conference on Parallel and distributed processing
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper demonstrates the one-sided communication used in languages like UPC can provide a significant performance advantage for bandwidth-limited applications. This is shown through communication microbenchmarks and a case-study of UPC and MPI implementations of the NAS FT benchmark. Our optimizations rely on aggressively overlapping communication with computation, alleviating bottlenecks that typically occur when communication is isolated in a single phase. The new algorithms send more and smaller messages, yet the one-sided versions achieve 1.9× speedup over the base Fortran/MPI. Our one-sided versions show an average 15% improvement over the twosided versions, due to the lower software overhead of onesided communication, whose semantics are fundamentally lighter-weight than message passing. Our UPC results use Berkeley UPC with GASNet and demonstrate the scalability of that system, with performance approaching 0.5 TFlop/s on the FT benchmark with 512 processors.