Optically Switchable Memristors: A Route to Hierarchical Control in Artificial Intelligent Systems

Abstract

Modulation of resistive switching memory by light opens the route to new optoelectronic devices that can be switched optically and read electronically. Applications include integrated circuits with memory elements switchable by light and optically reconfigurable and tunable synaptic circuits for neuromorphic computing applications and brain-inspired artificial intelligent systems. A novel polarization and wavelength-specific optoelectronic memory device is presented1 that can be controlled purely optically, electronically or by both. Optical or electronic modulation switches the device between low and high conducting states whilst modulation by both facilitates fine-tuning of the resistive memory properties and switching characteristics of the device. In biological synapses, the connection strength (plasticity) between two neurons is controlled by the ionic flow through the synaptic cleft and it is widely believed that the adaptation of synaptic weights enables biological systems to learn and function. Similarly, the conductance of a memristor depends on the history of the total charge that has travelled through it. A key feature of neuronal learning is habituation, whereby repeated stimuli strengthens the synaptic plasticity whilst a lack of stimuli results in weakening. Learning in biological systems also involves spike-timing-dependent plasticity (STDP). In STDP, learning the synaptic efficacy governing potentiation and depression is determined by the temporal order of pre-synaptic and post-synaptic spikes. Using our optical memristor device, we demonstrate optical control of synaptic potentiation and depression, optical switching between short and long-term memory and optical modulation of the synaptic efficacy via spike timing dependent plasticity. The work opens the route to dynamic patterning of memristor networks both spatially and temporally by light, thus allowing the development of new optically reconfigurable neural networks and adaptive electronic circuits.2 , 3 References Jaafar, A. H. et al. Reversible optical switching memristors with tunable STDP synaptic plasticity: a route to hierarchical control in artificial intelligent systems. Nanoscale 9, 17091 (2017).Kossifos, K. M., Antoniades, M. A., Georgiou, J., Jaafar, A. H. & Kemp, N. T. An Optically-Programmable Absorbing Metasurface. IEEE Int. Symp. Circuits Syst. (2018). doi:10.1109/ISCAS.2018.8351874Georgiou, J., Kossifos, K. M., Antoniades, M. A., Jaafar, A. H. & Kemp, N. T. Chua Mem-Components for Adaptive RF Metamaterials. IEEE Int. Symp. Circuits Syst. (2018). doi:10.1109/ISCAS.2018.8351852 Figure 1