Xiangxiang Gao, Hong Zhang, Yuelong Feng, Jian Li, Yufeng Chen, Miao Zhang, Zhenhua Lin, Jincheng Zhang, Yue Hao, Jingjing Chang
ABSTRACT Driven by artificial intelligence (AI) and Internet of Things (IoT) technologies, the “memory wall” and “power wall” issues of the traditional von Neumann architecture have become increasingly prominent. Neuromorphic computing has emerged as a crucial breakthrough direction due to its advantages of imitating the human brain's in‐memory computing and low power consumption, with synaptic devices being its core component. As a wide‐bandgap semiconductor, gallium oxide (Ga 2 O 3 ) possesses excellent thermal stability, chemical inertness, and radiation resistance. Moreover, its abundant defect states (e.g., oxygen vacancies) can be dynamically modulated via light and electrical stimuli, enabling Ga 2 O 3 ‐based devices to exhibit synaptic‐like behaviors, thus making it an ideal material for simulating biological synapses. This review systematically examines cutting‐edge developments in Ga 2 O 3 ‐based optoelectronic devices for neuromorphic computing applications. First, this paper provides a comprehensive overview of the behaviors of biological synapses. Subsequently, it introduces the operating principles and core device performance metrics of Ga 2 O 3 ‐based memristors, phototransistors, and photodetectors for neuromorphic computing. Finally, the synaptic plasticity of Ga 2 O 3 optoelectronic devices and their applications in neuromorphic computing were summarized, providing a material‐device‐system full‐architecture solution for the next generation of neuromorphic chips.