Parallel inference machine PIM:

Abstract: "The parallel inference machine, PIM, is the prototype hardware system in the Fifth Generation Computer System (FGCS) project. The PIM system aims at establishing the basic technologies for large-scale parallel machine architecture, efficient kernel language implementation and many as...

Full description

Saved in:
Bibliographic Details
Main Author: Taki, Kazuo (Author)
Format: Book
Language:English
Published: Tokyo, Japan 1992
Series:Shin-Sedai-Konpyūta-Gijutsu-Kaihatsu-Kikō <Tōkyō>: ICOT technical report 775
Subjects:
Summary:Abstract: "The parallel inference machine, PIM, is the prototype hardware system in the Fifth Generation Computer System (FGCS) project. The PIM system aims at establishing the basic technologies for large-scale parallel machine architecture, efficient kernel language implementation and many aspects of parallel software, that must be required for high performance knowledge information processing in the 21st century. The PIM system also supports an R & D environment for parallel software, which must extract the full power of the PIM hardware. The parallel inference machine PIM is a large-scale parallel machine with a distributed memory structure. The PIM is designed to execute a concurrent logic programming language very efficiently
The features of the concurrent logic language, its implementation, and the machine architecture are suitable not only for knowledge processing, but also for more general large problems that arise [sic] dynamic and non-uniform computation. Those problems have not been covered by commercial parallel machines and their software systems targeting scientific computation. The PIM system focuses on this new domain of parallel processing. There are two purposes to this paper. One is to report an overview of the research and development of the PIM hardware and its language system
The other is to clarify and itemize the features and advantages of the language, its implementation and the hardware structure with the view that the features are strong and indispensable for efficient parallel processing of large problems with dynamic and non-uniform computation.
Physical Description:23 S. graph. Darst.

There is no print copy available.

Interlibrary loan Place Request Caution: Not in THWS collection!