Memristor‐Based Neuromorphic Computing for Intelligent Sensing and Edge Applications

ABSTRACT

Memristor‐based neuromorphic computing has emerged as a promising hardware paradigm for energy‐efficient, scalable artificial intelligence edge applications. This review surveys recent progress in memristor‐based neuromorphic computing, in‐memory computing architectures, learning algorithms, and hardware implementations for edge intelligence. Volatile and nonvolatile memristive dynamics are discussed as key mechanisms for emulating biological neurons and synapses, providing foundational building blocks for crossbar‐based hardware neural networks. Moving beyond conventional neural‐network acceleration, this review examines the integration of memristive technologies into edge‐specific applications, focusing on emerging in‐sensor computing, optical in‐memory computing, and multimodal sensing frameworks. Furthermore, we outline the deployment of hardware‐compatible, brain‐inspired algorithms—including artificial, spiking, and recurrent neural networks—and evaluate their physical implementation within memristive architectures. The review also highlights the evolution of memristor‐based hardware from single‐device demonstrations to array‐level accelerators and CMOS‐integrated system‐on‐chip platforms, with the associated hierarchical challenges and solutions. By bridging the gap between individual device physics and full system engineering, this review provides a comprehensive roadmap for developing highly adaptive, low‐latency, and autonomous hardware platforms for the next generation of localized edge intelligence.