Recent progress in brain-like chips based on optoelectronic memory: Structures, mechanisms, and prospects

Abstract

Optoelectronic brain-like devices have been widely investigated in recent years and are considered a new generation of hardware platforms for neuromorphic computing. Inspired by the biological visual perception system, the devices integrate sensing, computing, and memory in a single functional unit. Compared with the electronically controlled memristor, the introduction of optical signals can further improve the computational efficiency and physically integrate the sensing unit and the processing unit. This efficient and intelligent information processing method can effectively overcome the bottleneck of traditional computing under von Neumann architecture. However, the development of optoelectronic brain-like devices is still in the preliminary stage, and its mechanism is complex and not uniform. Therefore, it is necessary to deeply understand the quantum process between optical input and electrical output to provide a better reference for the development of this field. This article aims to comprehensively review the latest progress in optoelectronic brain-like devices, summarizing the device performance and structures. It also provides a comprehensive summary of multiple mechanisms under different material systems, such as direct photoelectric conversion or photoelectric conversion triggering subsequent effects. In addition, a variety of potential application scenarios for optoelectronic devices are introduced. Finally, we present some possible problems in the development of this field. This review can help researchers better understand the whole picture of the development of optoelectronic devices.