神经形态工程学
记忆电阻器
计算机科学
材料科学
人工神经网络
光电子学
人工智能
电子工程
工程类
作者
Chunmei Li,Yanan Li,Haolin Luo,Yangming Leng,Jinyong Wang,Wei Li
摘要
With the rapid development of intelligent era, machine vision based on von Neumann architecture will face a bottleneck, while the developing neuromorphic computing is one of the most promising options to solve the bottleneck. The key step is to enable the artificial devices to simulate the behaviors of biological synapses, and light-modulated memristors can be applied in neuromorphic computing to build artificial visual systems. Here, a bilayer memristive film of SrTiO3 /CuAlO2 was chosen to construct a two-terminal artificial synaptic sample with construction of ITO/SrTiO3:Ru/CuAlO2/ITO, in which the SrTiO3 layer was doped by Ruthenium (Ru) in order to enhance the light response from ultraviolet to visible range. It is found that the number of interior carries in the memristor could be generated by exterior stimulation of light, and the fabricated memristor could successfully mimic multiple synaptic functions of the biological synapse like excitatory postsynaptic current, short/long-term plasticity transform, paired-pulse facilitation and so on. The obtained excellent synaptic plasticities and optoelectronic properties of our light-modulated memristor have shown a potential in the machine vision and artificial neuromorphic computing systems in the near future.
科研通智能强力驱动
Strongly Powered by AbleSci AI