Abstract
Data structures such as sets, relations and graphs occur frequently in Artificial Intelligence applications. When these structures are large, as in deductive databases and knowledge bases, the performance of conventional computers is poor. The generic operations on these structures, such as pattern-directed search, set intersection and transitive closure, all involve associative matching of variable-length lexical strings. In this paper, which is an updated version of a recent UNICOM presentation, we describe novel add-on hardware which employs SIMD parallel techniques to achieve the required associative processing. The hardware is embedded in an architecture which allows knowledge bases to be stored and processed in situ. Performance figures are given.
This is a preview of subscription content, log in via an institution to check access.
Preview
Unable to display preview. Download preview PDF.
Similar content being viewed by others
References
R B K Dewar,E Schonberg and J T Schwartz ‘High-Level Programming — An Introduction to the Programming Language SETL’ Courant Institue of Math Sciences, New York 1983.
S H Lavington, ‘Technical Overview of the Intelligent File Store'. Knowledge-Based Systems, Vol.1, No.3, June 1988, pages 166–172.
S H Lavington and R A J Davies ‘Active Memory for Managing Persistent Objects'. Presented at the International Workshop on Computer Architectures to Support Security and Persistence, Bremen, FRG, May 1990. Proceedings to be published by Springer-Verlag.
J Rosenberg and J L Keedy 'software Management of a Large Virtual Memory'. Proceedings of the 4th. Australian Computer Science Conference, Brisbane, February 1981, pages 171–183.
J Robinson and S H Lavington ‘A Transitive Closure and Magic Functions Machine’ To be presented at the Second International Symposium on Databases in Parallel and Distributed Systems, July 1990, Dublin.
C Freuder ‘Synthesizing Constraint Expressions’ CACM November 1978, Vol 21, No 11, 958–966.
J De Kleer ‘An Assumption-Based TMS’ AI 28 (1986) 127–162.
C J Wang and S H Lavington 'sIMD Parallelism for Symbol Mapping'. To be presented at the International Workshop on VLSI for AI and Neural Networks, Oxford, September 1990.
S H Lavington, J Robinson and K Y Mok, ‘A High Performance Relational Algebraic Processor for Large Knowledge Bases'. Presented at the International Workshop on VLSI for Artificial Intelligence, Oxford, July 1988. Published in: VLSI for Artificial Intelligence, eds. Delgado-Frias and Moore, Kluwer Academic Press, 1989, pages 133–143.
S H Lavington, M Standring, Y J Jiang, C J Wang and M E Waite, ‘Hardware Memory Management for Large Knowledge Bases'. Proceedings of PARLE, the conference on Parallel Architectures and Languages Europe, Eindhoven, June 1987, pages 226–241. (Published by Springer-Verlag as Lecture Notes in Computer Science, Nos. 258 & 259).
H Walther ‘Performance Measurement of the Associative Memory IFS at the Artificial Intelligence Applications Institute at Edinburgh, 13 November 1989–24 November 1989'. UBILAB Report February 1989, Union Bank of Switzerland, Zurich.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1991 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Lavington, S. (1991). Parallel associative processing for knowledge bases. In: America, P. (eds) Parallel Database Systems. PDS 1990. Lecture Notes in Computer Science, vol 503. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-54132-2_53
Download citation
DOI: https://doi.org/10.1007/3-540-54132-2_53
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-54132-5
Online ISBN: 978-3-540-47432-6
eBook Packages: Springer Book Archive