Content-Addressable Memory Manager: Design and Evaluation.

Abstract

This research covers basic issues in implementing a Content-Addressable Memory (CAM) for a microcomputer. The thesis compares the various types of CAM organizations and presents a set of standards to distinguish true CAM from CAM-like memory. Issues such as the meaning of 'content' to the host computer are discussed. This discussion leads to the implementation issues of how the CAM should operate with the host computer. The two configurations addressed in detail: a completely integrated approach and a peripheral approach. The thesis does not address the associative processor configurations found in PEPE, STARAN, or MPP. The case is presented that the peripheral device approach is the most practical approach. The peripheral device, CAM board, using the AFIT CAM IC chips, will be controlled from a dedicated device that is designed and evaluated in this thesis. The architecture and operation of the CAM IC chip controller (CAM Manager) is derived from the desire to have a general purpose device that will have a small instruction set. The instructions will allow for efficient operations on applications as diverse as LISP garbage collection, database searching, and array arithmetic. This thesis compares a conventional computer executing the test applications and the same computer using the CAM board. The test applications were: (a) addition over all elements of an array, (b) LISP garbage collection, and (c) image recognition. The CAM board can enhance the performance of the host microcomputer. This initial implementation is not suited for large databases. This restriction was due to the size of the board's content-addressable memory (32K bits total).

Document Details

Document Type
Technical Report
Publication Date
Dec 01, 1985
Accession Number
ADA164037

Entities

People

  • Mark C. Rowe

Organizations

  • Air Force Institute of Technology

Tags

DTIC Thesaurus Topics

  • Arithmetic
  • Computers
  • Computing Devices
  • Computing System Architectures
  • Content Addressable Memory
  • Databases
  • Host Computers
  • Image Recognition
  • Instruction Set Architecture
  • Instructions
  • Personal Computers
  • Recognition
  • Standards
  • Test And Evaluation

Fields of Study

  • Computer science

Readers

  • Aerospace Engineering
  • Parallel and Distributed Computing.

Technology Areas

  • AI & ML
  • AI & ML - DoD AI Strategy