Organic and perovskite memristors have superior characteristics both in material and structural perspectives, and therefore have been evaluated for possible integration as bio-realistic components of artificial intelligent hardware systems. This application will require the brain-inspired integrated systems that can process and memorize large amounts of complex information; requirements include highly uniform and reliable memristors that can be operated at low energy and integrated at high density. Here, we review the progress in development of organic and perovskite memristors to obtain various synaptic behaviors, with focus on material and underlying mechanism aspects. Then we address various approaches to meet the needs for constructing applications of neuromorphic computing, including low energy consumption, high uniformity and reliability of the memristors, and high-density integration. Lastly, we suggest future research directions toward realizing neuromorphic computing.

有机和钙钛矿忆阻器在材料和结构方面都具有优越的特性，因此已被评估为可能集成为人工智能硬件系统的生物现实组件。该应用程序将需要能够处理和记忆大量复杂信息的受大脑启发的集成系统；要求包括高度均匀和可靠的忆阻器，它们可以在低能量下运行并以高密度集成。在这里，我们回顾了有机和钙钛矿忆阻器的开发进展，以获得各种突触行为，重点是材料和潜在机制方面。然后我们提出了各种方法来满足构建神经形态计算应用程序的需求，包括低能耗、忆阻器的高均匀性和可靠性，以及高密度集成。最后，我们提出了实现神经形态计算的未来研究方向。