实现(概率)
记忆电阻器
铁电性
兴奋剂
材料科学
功率(物理)
薄膜
光电子学
电子工程
电气工程
纳米技术
物理
工程类
数学
量子力学
电介质
统计
作者
Xiaobing Yan,Jiahao Bai,Yinxing Zhang,Hong Wang,Jianhui Zhao,Zhenyu Zhou,Yong Sun,Zhongrong Wang,Zhenqiang Guo,Zhen Zhao,Jiangzhen Niu
标识
DOI:10.1016/j.mtnano.2024.100458
摘要
Next-generation non-volatile memories with low power consumption and multiple functions are highly desired in the era of big data. Here, we report a high-performance ferroelectric memristor Pd/Lu:HfO2/La0.7Sr0.3MnO3/SrTiO3/Si, which has low average set and reset powers of 4.28 nW and 0.75 nW, respectively. It can successfully simulate biological synaptic functions, such as the refractory period, spike-timing-dependent plasticity, and paired pulse facilitation. Moreover, the device realizes the simulation of decimal addition and multiplication. In particular, the device can be combined with the threshold device to form a neuromorphic network, which not only realizes multi-value data storage, data decryption and encryption, but also intensity modulation of peak frequency. These results may throw light on the way for the future research of low-power multifunctional neuro morphology chips.
科研通智能强力驱动
Strongly Powered by AbleSci AI