A Survey of Analog Memory Devices

Abstract

Widespread and persistent interest in the impleunit. Typical figures derived from computer simulation mentation of multilevel logic, conditional probability computers, range from 30 to 300 levels.'-3 learning machines, and brain models has created a need for an in-W. . b expensive analog or quasi-digital storage element. A number of possible approaches to this problem, ranging from the slow and the value one level at a time, it is not necessary to be reliable electromechanical systems to the many forms of charge and able to reproduce a given level exactly, as it would be if flux integration, are reviewed, and the suitability of each device for the machine were to perform arithmetic operations. A various fields of application is briefly discussed. way of resetting the value to some arbitrary zero level

Multi-level memristive memory with resistive networks

“…One of them is the evolution of many valued logic gates that push the speed limits of computing [4]. The other motivation of designing analog memory is that it can be used in neuromorphic designs, as it is useful for synaptic-weight storage in artificial neural networks [5]. One of the early designs that exploited analog memory in neural networks was Adaline (Adaptive Linear Neuron or later Adaptive Linear Element) which memory element consisted of memistors [6].…”

Unconventional computing

“…Notable among other attempts to achieve the integration and storage needed for adaptive systems was the "memistor" of Widrow (1962) which utilised electrolytic deposition and removal of metal onto a carbon filament whose resistance was thereby altered. An early survey of such schemes was by Nagy (1963). An obvious means of integration and analog storage is by charge on a capacitor, for example in the conditional probability computer of Uttley (Andrew, 1959), where slow leakage is inevitable but can be said to model forgetting in biological memory.…”