Nov. 15, 2003 to Nov. 21, 2003
D. Brent Weatherly , The University of Georgia, Athens
David K. Lowenthal , The University of Georgia, Athens
Mario Nakazawa , The University of Georgia, Athens
Franklin Lowenthal , California State University, Hayward
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/SC.2003.10016
Distributing data is a fundamental problem in implementing efficient distributed-memory parallel programs. The problem becomes more difficult in environments where the participating nodes are not dedicated to a parallel application. We are investigating the data distribution problem in non dedicated environments in the context of explicit message-passing programs.<div></div> To address this problem, we have designed and implemented an extension to MPI called Dynamic MPI (Dyn-MPI). The key component of Dyn-MPI is its run-time system, which efficiently and automatically redistributes data on the fly when there are changes in the application or the underlying environment. Dyn-MPI supports efficient memory allocation, precise measurement of system load and computation time, and node removal. Performance results show that programs that use Dyn-MPI execute efficiently in non dedicated environments, including up to almost a three-fold improvement compared to programs that do not redistribute data and a 25% improvement over standard adaptive load balancing techniques.
D. Brent Weatherly, David K. Lowenthal, Mario Nakazawa, Franklin Lowenthal, "Dyn-MPI: Supporting MPI on Non Dedicated Clusters", SC, 2003, SC Conference, SC Conference 2003, pp. 5, doi:10.1109/SC.2003.10016