Mechanism-Guided Design Rules for Neuromorphic Resistive Switching Devices

Abstract Resistive switching devices have emerged as promising candidates for neuromorphic comput ing due to their scalability, low power consumption, and inherent ability to emulate synaptic plasticity. Despite rapid progress, the field still lacks a unified framework that systematically links fundamental switching mechanisms with practical device design, reliability challenges, and pathways toward commercialization. In this review, we present a mechanism-driven decision making framework for the design of neuromorphic resistive switching devices. The approach establishes clear connections between switching physics and key performance trade-offs, failure mechanisms, and sustainability considerations. A comprehensive analysis of 293 studies pub lished between 2018 and 2026 is carried out to evaluate four primary switching mechanisms: valence change memory, electrochemical metallization, thermochemical switching, and ferro electric switching. By correlating material properties, device architectures, and operational characteristics, this work provides practical design guidelines for optimizing device performance while addressing reliability and scalability concerns. The proposed framework aims to bridge the gap between fundamental understanding and real-world implementation, thereby accelerating the development of next-generation neuromorphic hardware