Processor Allocation and Scheduling of Macro Dataflow Graphs on Distributed Memory Multicomputers by the PARADIGM Compiler
- 1 August 1993
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 134-138
- https://doi.org/10.1109/icpp.1993.153
Abstract
Functional or Control parallelism is an efiectiuc way to increase speedups in Multicomputers. Programs for these machines are represented by Macro Dataflow Graphs (MRGsJ for the purpose of functional parallelism analysk and exploitation. Algorithms for allocation and schedttlang of MDGs have been discussed along with some analysis of their optirnality. These algorithms attempt to minimize the execution time of any given MDG through exploitation of functional parallelism. Our preliminary results show their eflectiveness over naive algorithms.Keywords
This publication has 9 references indexed in Scilit:
- A parallel programming tool for scheduling on distributed memory multiprocessorsPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2003
- A scheduling algorithm for parallelizable dependent tasksPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- Approximate algorithms scheduling parallelizable tasksPublished by Association for Computing Machinery (ACM) ,1992
- A Heuristic of Scheduling Parallel Tasks and Its AnalysisSIAM Journal on Computing, 1992
- Demonstration of automatic data partitioning techniques for parallelizing compilers on multicomputersIEEE Transactions on Parallel and Distributed Systems, 1992
- Automatic extraction of functional parallelism from ordinary programsIEEE Transactions on Parallel and Distributed Systems, 1992
- A fast static scheduling algorithm for DAGs on an unbounded number of processorsPublished by Association for Computing Machinery (ACM) ,1991
- Performance Guarantees for Scheduling AlgorithmsOperations Research, 1978
- Complexity of Scheduling under Precedence ConstraintsOperations Research, 1978