神经形态工程学
磁畴壁(磁性)
材料科学
自旋电子学
领域(数学分析)
记忆电阻器
人工神经网络
计算机数据存储
凝聚态物理
磁化
计算机科学
光电子学
纳米技术
物理
电子工程
铁磁性
磁场
人工智能
数学
工程类
操作系统
数学分析
量子力学
作者
Bo Zhang,Yunzhuo Wu,Xiaoxue Zeng,Yonghai Guo,Bo Wang,Jiangwei Cao
摘要
Spintronic devices can realize multilevel state storage and mimic the properties of the synapse, which enables their potential application in the field of artificial neural networks. In this paper, we demonstrate the existence of a large intermediate transition zone in current-induced magnetization switching curves of Pt/Co–Tb/Ta structures, and the number of states in the transition zone that can be manipulated by changing the Co content. The magneto-optical Kerr microscope imaging indicates that this property is related to the constrained domain wall motion in the Co–Tb films with large Co content. We also demonstrate the multilevel state storage properties of the sample by applying a sequence of current pulses. The synaptic plasticity behaviors were mimicked in these samples through regulating the value of Hall resistance by current pulses. The constrained domain wall motion supplies a simple but effective way to achieve multilevel state storage and show potential applications in neuromorphic computing.
科研通智能强力驱动
Strongly Powered by AbleSci AI