Optoelectronic Memristors (OMs) represent a significant hardware foundation for constructing artificial visual neural networks. As a novel class of integrated sensory-memory-computing devices, they hold great promise for overcoming the bottlenecks inherent in traditional von Neumann computing architectures. Leveraging desirable characteristics such as high bandwidth and low power consumption, OMs integrate optical sensing, information storage, and neuromorphic computing functionalities. This integration endows them with substantial potential for brain-inspired visual neural systems. This review summarizes recent progress in OMs, focusing on materials and physical mechanisms, performance metrics, and multi-mode in-sensor computing applications. The applications of oxides, two-dimensional materials, chalcogenides, and biomaterials in OMs are detailed, with corresponding operating mechanisms analyzed. Subsequently, the fundamental electrical properties and optoelectronic response characteristics of OMs are analyzed. Furthermore, synaptic plasticity in OMs is discussed, encompassing short-term/long-term plasticity learning rules and other neuromorphic functionalities emulation, based on their inherent neuromorphic properties. Additionally, applications of OMs in Boolean logic operations, artificial vision systems, and wearable neuromorphic devices are examined. Conclusively, the primary advantages, persistent challenges, and emerging research trajectories of OMs are synthesized. This analysis establishes foundational insights for advancing brain-inspired neural systems.
扫码关注我们
求助内容:
应助结果提醒方式:
