Chapter

Euro-Par 2012 Parallel Processing

Volume 7484 of the series Lecture Notes in Computer Science pp 28-39

Pattern-Independent Detection of Manual Collectives in MPI Programs

  • Alexandru CalotoiuAffiliated withGerman Research School for Simulation SciencesComputer Science Department, RWTH Aachen University
  • , Christian SiebertAffiliated withGerman Research School for Simulation SciencesComputer Science Department, RWTH Aachen University
  • , Felix WolfAffiliated withGerman Research School for Simulation SciencesComputer Science Department, RWTH Aachen UniversityForschungszentrum Jülich, Jülich Supercomputing Centre

* Final gross prices may vary according to local VAT.

Get Access

Abstract

In parallel applications, a significant amount of communication occurs in a collective fashion to perform, for example, broadcasts, reductions, or complete exchanges. Although the MPI standard defines many convenience functions for this purpose, which not only improve code readability and maintenance but are usually also highly efficient, many application programmers still create their own, manual implementations using point-to-point communication. We show how instances of such hand-crafted collectives can be automatically detected. Matching pre- and post-conditions of hashed message exchanges recorded in event traces, our method is independent of the specific communication pattern employed. We demonstrate that replacing detected broadcasts in the HPL benchmark can yield significant performance improvements.

Keywords

MPI collective operations performance optimization HPL