Dynamical Symbiosis of Solar Cell and Memristor

Memristive devices have attracted significant attention due to their nonlinear dynamics, analog tunability, and ability to emulate synaptic functions. When combined with energy-harvesting components, memristors offer an opportunity to realize self-powered physical AI systems. We demonstrate a solar cell-driven perovskite memristor architecture, which is suitable for autonomous physical AI platforms for light-driven computation. In this system, incident light is directly converted into electrical stimuli by integrated solar cells, which modulate the conductance states of memristors. These conductance changes can encode computational states and learning behaviors, enabling direct processing of optical information without intermediate digital conversion. Such a light-driven memristive system enables simultaneous sensing, energy harvesting, memory storage, and computation within a unified physical structure.