Optimization Rules for Programming with Collective Operations

  title={Optimization Rules for Programming with Collective Operations},
  author={Sergei Gorlatch and Christoph Wedler and Christian Lengauer},
We study how several collective operationslike broadcast, reduction, scan, etc. can be composed efficiently in complex parallel programs. Our specific contributions are: (1) a formal framework for reasoning about collective operations; (2) a set of optimization rules which save communications by fusing several collective operations into one; (3) performance estimates, which guide the application of optimization rules depending on the machine characteristics; (4) a simple case study with machine… CONTINUE READING


Publications citing this paper.
Showing 1-10 of 32 extracted citations


Publications referenced by this paper.
Showing 1-10 of 21 references

A ZPL programming guide (Version 6.2)

  • L. Snyder
  • Technical report, Department of Computer Science…
  • 1998
Highly Influential
3 Excerpts

NESL: A nested data-parallel language

  • G. Blelloch
  • Technical Report CMU-CS-93-129, Department of…
  • 1992
Highly Influential
5 Excerpts

Homomorphism (de)compositions for parallel programming

  • S. Gorlatch, C. Lengauer
  • In Proc. 3rd Int. Working Conf. on Massively…
  • 1998
1 Excerpt

Optimizing compositions of scans and reductions in parallel program derivation

  • S. Gorlatch
  • Technical Report MIP-9711, Universität Passau…
  • 1997
1 Excerpt

Similar Papers

Loading similar papers…