TITLE: Towards the Visual Microprocessor: VLSI Design and the Use of
Cellular Neural Network Universal Machines
AUTHOR: Tamás Roska
Visual microprocessors are software-controlled analogic (analog-and-logic)
array computers on a single chip. To create them requires designing highly
complex integrated circuits called VLSIs (Very Large Scale Integration)
operating in mixed-mode (analog and digital). The architecture of these
chips is based on Cellular Neural/nonlinear Networks (CNNs) extended
by building blocks to become a full-featured array computer, called the
CNN Universal Machine (CNN-UM). The elementary component-computers
are electronic mini organism-like cells that can be re-programmed either
externally or internally.
Opening with a foreword by the respected academic, Professor Leon Chua,
the book progresses to explore circuit design, prototyping and analogial
algorithms. Subjects covered include the VLSI design and implementation
of CNNs and CNN-UMs, the testing of CNN chips and a detailed analysis
of a new chip prototyping system and interfacing the CNN universal chips
to a digital environment.
Distinctive features include:
- a textbook-like introduction to the new field of analogic super-computing
- design methodology of CMOS mixed-mode (analogic) array computers
and optical sensors
- applications in neuro-computing, machine vision, image processing
and VSL signal processing
- design methodology of a new type of algorithm for solving complex
- important pioneering work about a rapidly-developing and increasingly
influential field of computation circuit design, and sensory computers
- a vertical coverage of topics on a new type of computer from high
level language down to physical circuit design
The wide spectrum covered in this volume will be beneficial for students
engaging in advanced research, professionals in electrical and electronic
engineering, computer engineering, computer science, mathematics and
TABLE OF CONTENTS:
List of Contributors.
The Cellular Neural Network (CNN) and the CNN Universal Machine: Concept,
Architecture and Operation Modes (L. Chua & T. Roska).
CNN Chip Prototyping and Development Systems (Á. Zarándy, et al.).
CMOS Analogue Design Primitives (Á. Rodríguez-Vázquez, et al.).
On the Implementation of Linear and Nonlinear Interaction Operators for CNNS
(Á. Rodríguez-Vázquez, et al.).
Light-Sensitive Devices in CMOS (E. Roca, et al.).
Examples of CMOS CNN Universal Chip Design (R. Domínguez-Castro, et al.).
Short-Term Storage of Analogue Signals for CNN Universal Machine Based
Image Processing (R. Carmona, et al.).
Application Case Studies (G. Liska, et al.).