Process-Oriented Collective Operations

Bjørndalen, John Markus and Sampson, Adam T. (2008) Process-Oriented Collective Operations. In: Welch, Peter H. and Stepney, Susan and Polack, Fiona A.C. and Barnes, Frederick R.M. and McEwan, Alistair A. and Stiles, Gardner S. and Broenink, Jan F. and Sampson, Adam T., eds. Communicating Process Architectures 2008. Concurrent Systems Engineering . IOS Press, Amsterdam, Netherlands, pp. 309-328. ISBN 978-1-58603-907-3. E-ISBN 978-1-60750-368-2. (doi:10.3233/978-1-58603-907-3-309)

Abstract

Distributing process-oriented programs across a cluster of machines requires careful attention to the effects of network latency. The MPI standard, widely used for cluster computation, defines a number of collective operations: efficient, reusable algorithms for performing operations among a group of machines in the cluster. In this paper, we describe our techniques for implementing MPI communication patterns in process-oriented languages, and how we have used them to implement collective operations in PyCSP and occam-pi on top of an asynchronous messaging framework. We show how to make use of collective operations in distributed processoriented applications. We also show how the process-oriented model can be used to increase concurrency in existing collective operation algorithms.

Item Type:	Book section
DOI/Identification number:	10.3233/978-1-58603-907-3-309
Uncontrolled keywords:	occam-pi python pycsp mpi distributed
Subjects:	Q Science > QA Mathematics (inc Computing science) > QA 76 Software, computer programming,
Divisions:	Faculties > Sciences > School of Computing > Systems Architecture Group
Depositing User:	Mark Wheadon
Date Deposited:	29 Mar 2010 12:14 UTC
Last Modified:	03 Jun 2019 09:13 UTC
Resource URI:	https://kar.kent.ac.uk/id/eprint/24102 (The current URI for this page, for reference purposes)