author : F. R. Pettit

The author commences with a survey of the development of electronic systems of the past 50 years and unravels some of the aspects: complexity, failure rate, repair time, utility of equipment and systems, and availability. After discussion of the nature of components and interconnections, and their failure modes, he cites instances of failures in large systems. He then examines the limits of complexity, and suggests where continued improvement can be made in the availability of systems which will go further towards satisfying their operational requirements.

Because the question of reliability presents serious problems with both analog and digital processors, a new form of processing is considered here, having the unique characteristic of being unaffected by failures of components and interconnections. The concept of a Pulsator Matrix is developed theoretically, and a number of potential matrix forms examined. Computer simulations are described, and the way pointed to full-scale development of the vast, complex processors in microelectronic form, with the interesting property that 100% production yield is possible, because defective parts are acceptable, desirable almost, in Pulsatory Matrix Processors.

The author asks the question: can we construct a processor which does not use numbers, which operates using thousands of associative parallel operations, and which is independent of the fault conditions which naturally develop in electronic equipment? in the search for such a device, computer programming techniques are examined which permit the computer to model systems whose complexity greatly exceeds the storage capacity of the computer.