My, and others', spiking memristors are true memristors: a response to R.S. Williams' question at the New Memory Paradigms: Memristive Phenomena and Neuromorphic Applications Faraday Discussion (1812.05976v1)

Abstract: At the Faraday Discussion, in the paper titled `Neuromorphic computation with spiking memristors: habituation, experimental instantiation of logic gates and a novel sequence-sensitive perceptron model' it was demonstrated that a large amount of computation could be done in a sequential way using memristor current spikes (d.c. response). As these spikes are found in many memristors (possibly all), this novel approach could be highly useful for fast and reproducible memristor circuits. However, questions were raised as to whether these spikes were actually due to memristance or merely capacitance in the circuit. In this longer version of the Faraday Discussion response, as much information as is available from both published and unpublished data from my lab is marshalled together. We find that the devices are likely imperfect memristors with some capacitance, and that the spikes are related to the frequency effect seen in memristor hysteresis curves, thus are an integral part of memristance.