1.Fundamentals Department, Air Force Engineering University, Xi’an 710051, China 2.College of Computer, National University of Defense, Changsha 410005, China 3.Airforce Command College, Beijing 100097, China
Fund Project:Project supported by the National Natural Science Foundation of China (Grant No. 11975311) and the Natural Science Basic Research Program of Shaanxi, China (Grant Nos. 2021JM-221, 2020JQ-470).
Received Date:01 April 2021
Accepted Date:16 June 2021
Available Online:08 October 2021
Published Online:20 October 2021
Abstract:The spin neuron is an emerging artificial neural device which has many advantages such as ultra-low power consumption, strong nonlinearity, and high integration. Besides, it has ability to remember and calculate at the same time. So it is seen as a suitable and excellent candidate for the new generation of neural network. In this paper, a spin neuron driven by magnetic field and strain is proposed. The micromagnetic model of the device is realized by using the OOMMF micromagnetic simulation software, and the numerical model of the device is also established by using the LLG equation. More importantly, a three-layer neural network is composed of spin neurons constructed respectively using three materials (Terfenol-D, FeGa, Ni). It is used to study the activation functions and the ability to recognize the MNIST handwritten datasets.c Results show that the spin neuron can successfully achieve the random magnetization switching to simulate the activation behavior of the biological neuron. Moreover, the results show that if the ranges of the inputting magnetic fields are different, the three materials' neurons can all reach the saturation accuracy. It is expected to replace the traditional CMOS neuron. And the overall power consumption of intelligent computing can be further reduced by using appropriate materials. If we input the magnetic fields in the same range, the recognition speed of the spin neuron made of Ni is the slowest in the three materials. The results can establish a theoretical foundation for the design and the applications of the new artificial neural networks and the intelligent circuits. Keywords:nanomagnet/ spin neuron/ magnetization switching/ neural network computing
图3展示了纳磁体在应力和磁场共同作用下的磁化过程. 该过程主要分为3个阶段, 第1阶段是在0 ns时, 纳磁体在30 MPa的应力作用下, 磁化方向由初始方向翻转90°至x轴负方向, 到达空态. 第2阶段是在1 ns时, 撤去应力, 施加10 mT磁场, 将继续翻转. 若磁场方向与初始磁化方向一致, 磁化向下翻转回到初态; 若方向相反, 磁化向上翻转达到另一稳态. 第3阶段是2 ns后, 撤去磁场, 纳磁体将始终位于稳定状态. 由于OOMMF软件仿真没有考虑热噪声的影响, 因此无法体现出该自旋神经元结构纳磁体随机翻转的特性. 但是通过施加不同方向的磁场, 对动态磁化过程图进行分析研究后, 发现纳磁体可以在磁场和应力共同作用下完成翻转. 图 3 动态磁化过程 (a) 磁化初始方向为–y; (b) 施加30 MPa应力, 磁化翻转90°; 撤去应力, 施加10 mT磁场; (c)方向相反, 磁化翻转180°; (d)方向相同, 磁化翻转0° Figure3. Magnetization process: (a) Initial direction of magnetization is –y; (b) a 30 MPa strain is applied and then 90° switching is achieved, removing strain and applying a 10 mT magnetic field; (c) 180° magnetization switching; (d) 0° magnetization switching.
通过OOMMF仿真, 还获得了该自旋神经元结构的磁化矢量变化曲线, 如图4所示. 结合纳磁体磁化矢量的参数表达式, 见(5)—(7)式, 可以对磁化过程做进一步分析. 由于x与y方向上的磁化矢量mx, my与方位角φ的大小有关, 因此其值随方位角的改变在不断发生变化. 0 ns时, 纳磁体磁化翻转至空态, 此时my由–1变化到0. 1 ns时, 纳磁体在只有磁场的作用下发生翻转. 当磁场方向沿y轴正方向, my从0变化到+1; 当磁场方向沿y轴负方向, my则由0变化至–1. mx与my的变化过程正交. 图 4 磁化矢量随时间变化曲线图 (a) 1 ns时, 施加磁场沿y轴正方向时的磁化矢量变化图; (b) 1 ns时, 施加磁场沿y轴负方向的磁化矢量变化图 Figure4. Magnetization vector with time: (a) When t = 1 ns, a magnetic field is applied in the direction along +y; (b) when t = 1 ns, a magnetic field is applied in the direction along –y.