Efficient algorithms for block-cyclic array redistribution between processor sets
SC '98 Proceedings of the 1998 ACM/IEEE conference on Supercomputing
Interprocessor Communication with Limited Memory
IEEE Transactions on Parallel and Distributed Systems
Data Redistribution and Remote Method Invocation in Parallel Component Architectures
IPDPS '05 Proceedings of the 19th IEEE International Parallel and Distributed Processing Symposium (IPDPS'05) - Papers - Volume 01
MADRE: The Memory-Aware Data Redistribution Engine
Proceedings of the 15th European PVM/MPI Users' Group Meeting on Recent Advances in Parallel Virtual Machine and Message Passing Interface
Madre: the Memory-Aware Data Redistribution Engine
International Journal of High Performance Computing Applications
An in-place algorithm for irregular all-to-all communication with limited memory
EuroMPI'10 Proceedings of the 17th European MPI users' group meeting conference on Recent advances in the message passing interface
Optimizing Memory Constrained Environments in Monte Carlo Nuclear Reactor Simulations
International Journal of High Performance Computing Applications
Hi-index | 0.00 |
Many memory-bound distributed applications require frequent redistribution of data. Pinar and Hendrickson investigated two families of memory-limited redistribution algorithms. The first family has many advantages, but fails on certain inputs, and, if not implemented carefully, may lead to an explosion in the number of local data copies. The second family eliminates the possibility of failure at the expense of considerable additional overhead. We carefully analyze these algorithms and develop a modified method that potentially combines advantages of each. The resulting algorithm has been implemented in MADRE and experiments reveal its performance to be superior to that of other MADRE algorithms in most cases.