PaCT 1999: Parallel Computing Technologies pp 487-488 | Cite as
An Expert System Approach to Data Distribution and Distribution Analysis
Abstract
Research into systems aimed at transforming existing legacy programs to an equivalent form for improved execution performance on multiprocessor systems has led to the realisation that the level of parallelization expertise needed is similar to that of a human expert. Data distribution is one of the major obstacles limiting pure automation in parallelization. A system which automatically provides effective data partitioning algorithms can be considered suitable in replacing the human expert [1]. The Fortport project provides a solution to this problem by presenting a suite of tools to apply AI technology in the parallelization of sequential codes. This approach is based on an underlying knowledge model to influence the transformation process.
References
- [1]S. Andel, B. M. Chapman and H. P. Zima, An Expert Advisor for Parallel Programming Environments and its Realization within the Framework of the Vienna Fortran Compilation System, 4th Workshop on Compilers for Parallel Computers, Delft, 1993.Google Scholar
- [2]P. J. P. McMullan, P. Milligan, P. P. Sage and P. H. Corr. A Knowledge Based Approach to the Parallelisation, Generation and Evaluation of Code for Execution on Parallel Architectures. IEEE Computer Society Press, ISBN 0-8186-7703-1, pp 58-63, 1997.Google Scholar
- [3]P. J. P. McMullan, P. Milligan and P. H. Corr. Knowledge Assisted Code Generation and Analysis. Lecture Notes in Computer Science 1225, Springer Verlag, pp 1030–1031, 1997.Google Scholar