Dept. of Computer Science and Engineering
Oregon Graduate Center
Some biological aspects of neural interactions are presented and used as a basis for a computational model in the development of a new type of computer architecture. A VLSI microarchitecture is proposed that efficiently implements the neural-based computing methods. An analysis of the microarchitecture is presented to show that it is feasible using currently available VLSI technology. The performance expectations of the proposed system are analyzed and compared to conventional computer systems executing similar algorithms. The proposed system is shown to have comparatively attractive performance and cost/performance ratio characteristics. Some discussion is given on system level characteristics including initialization and learning.
Geller, Ronnie Dee, "A VLSI architecture for a neurocomputer using higher-order predicates" (1987). Scholar Archive. 234.