The Cellular Nonlinear Network (CNN) is a dynamic array of cells, each
of them has analog state, input and output. The mainly local
interactions between the cells in a spatial grid are characterized by a
cloning template or gene, specifying a complex spatial temporal operator
on a flow, as the elementary data. A flow on a 2D grid can be considered
as a video clip. Nonlinear wave operators are special cases (PDE
solvers)
The CNN Universal Machine is a stored programmable array computer on
flows. It is an analogic computer, analog wave dynamics are combined by
logic, locally and globally.
The CNN Universal Machine defines a new world of algorithms.
The CMOS implementation of the CNN Universal Machine on CMOS run with a
few TeraOPS on a sigle chip, in some cases the optical sensors are also
integrated. These Visual Microprocessors are used in many Mission
critical applications.
For molecular and nano device arrays, these Cellular Wave Computers
offer a very natural setting. In addition, the visual pathway and other
sensory organs are efficiently modelled ina programmable way.
Recent advanced image processing methods, like PDE based techniques,
level set and active waves, etc., inefficient on digital machines,are
extremely efficient in these machines.
The complexity theory of these machines will also be presented with live
demos of some applications.
Copyright. All Rights Reserved.