ParaGraph-RTS: A parallel reduction-based run-time system
We are investigating the use of the functional programming paradigm on multiprocessor computers. We have developed a parallel graph reduction run-time system called ParaGraph-RTS (Parallel Graph Reduction) that provides both significant amounts of exploitable parallelism and the robustness to support development of large scientific codes. This system allows us to empirically investigate the functional programming paradigm within computationally intensive application domains. In this paper, we describe an implementation of ParaGraph-RTS on a shared-memory multiprocessor. The implementation strategy reflects insights gained through performance characterizations of the reduction process. Preliminary performance results and analyses are presented. 8 refs., 4 figs.
- Research Organization:
- Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)
- DOE Contract Number:
- W-7405-ENG-36
- OSTI ID:
- 5956895
- Report Number(s):
- LA-UR-88-4110; CONF-890834-1; ON: DE89005457
- Resource Relation:
- Conference: International conference on parallel processing, St. Charles, IL, USA, 8-12 Aug 1989; Other Information: Portions of this document are illegible in microfiche products
- Country of Publication:
- United States
- Language:
- English
Similar Records
An efficient UNICOS run-time system for functional programs on a CRAY X-MP
Parallel String Graph Construction and Transitive Reduction for De Novo Genome Assembly