Much effort has been expended on developing special architectures dedicated to the efficient execution of production systems. While data-flow principles of execution offer the promise of high programmability for numerical computations, it is shown that data-driven principles can also be applied to symbolic computations. In particular, a mapping of the Rete match algorithm on the MIT tagged token data-flow architecture is considered as a benchmark. The results of a deterministic simulation of this multiprocessor architecture demonstrate that artificial intelligence production systems can be efficiently mapped on data-driven architectures.
|Number of pages
|Proceedings of the International Conference on Parallel Processing
|Published - Dec 1 1988
All Science Journal Classification (ASJC) codes
- Hardware and Architecture