空中骑兵
神经形态工程学
各向异性
物理
凝聚态物理
记忆电阻器
能量(信号处理)
铁磁性
计算机科学
光学
人工神经网络
量子力学
机器学习
作者
Ravish Kumar Raj,Ravi Shankar Verma,Brajesh Kumar Kaushik,Shailendra Yadav
摘要
The magnetic skyrmion has distinct features like nanoscale size, particle-like behavior, low driving current, and topologically stable which makes it a suitable candidate for neuromorphic computing. Synthetic antiferromagnetic (SAF) skyrmions consist of a pair of coupled ferromagnetic (FM) skyrmions, each in its respective sub-layers that are favourable over the FM skyrmions as they follow the straight trajectories and prevent its annihilation at the nanotrack edge. In this work, a leaky integrate and fire neuronal device model is proposed based on SAF skyrmions with voltage control magnetic anisotropy (VCMA) as a leaky effect for the tunability of the device. The anisotropy is directly correlated with the size of the skyrmion meaning that in the region with larger anisotropy, the skyrmion size is smaller and hence, more energy. However, the skyrmions have the tendency to move toward the minimum energy state means it will move towards the lower anisotropy. This behavior of SAF skyrmion on a nanotrack with anisotropy gradient corresponds to the leaky-integrate-fire (LIF) functionality of the neuron device. Moreover, device performance is also realized at room temperature for practical implementation. Hence, the proposed device possesses an energy-efficient artificial neuron opens up the path for the development of next-generation skyrmionic devices for neuromorphic computing.
科研通智能强力驱动
Strongly Powered by AbleSci AI