Cellular neural networks (CNNs) are a revolutionary new computing paradigm that allows for analog computation. CNNs consist of an array of identical cells arranged in a grid that are locally connected to neighboring cells. Each cell has input, output, and state variables and its output is related to its state through a nonlinear equation. The CNN universal machine (CNN-UM) was the first programmable analog processor based on a CNN that had its own programming language and operating system. CNNs have applications in areas like image processing, target recognition, and visual inspection.