Engineering the kinetics of redox-based memristive devices for neuromorphic computing

For neuromorphic computing applications, analog programming characteristics with linear updates are highly desirable. Here, we discuss the relation between the switching dynamics of redox-based memristive devices and their analog programming capability. We identity three groups of devices based on the valence change mechanism with characteristic properties. Experimental results and theoretical considerations suggest that analog switching can be achieved if a thermal runaway is avoided. Moreover, we argue that internal series resistances play a crucial role in controlling the runaway and determining the accessible resistance window.