Shape‐Dependent Multi‐Weight Magnetic Artificial Synapses for Neuromorphic Computing
Shape‐Dependent Multi‐Weight Magnetic Artificial Synapses for Neuromorphic Computing
复制标题
DOI：
10.1002/aelm.202200563
复制
发表时间：
2022-12
影响因子：
6.2
通讯作者：
Thomas Leonard;Samuel Liu;Mahshid Alamdar;Harrison Jin;Can Cui;Otitoaleke G. Akinola;Lin Xue;T. Xiao;J. Friedman;M. Marinella;C. Bennett;J. Incorvia
中科院分区：
文献类型：
--
作者：
Thomas Leonard;Samuel Liu;Mahshid Alamdar;Harrison Jin;Can Cui;Otitoaleke G. Akinola;Lin Xue;T. Xiao;J. Friedman;M. Marinella;C. Bennett;J. Incorvia
在神经形态计算中，人工突触提供一种多权重（MW）电导状态，该状态是根据来自神经元的输入设定的，类似于大脑。在此，对基于磁性材料且使用磁性隧道结（MTJ）和磁畴壁（DW）的人工突触进行了探究。通过在单个MTJ下方的DW轨道中制造光刻缺口，实现了3 - 5种稳定的电阻状态，这些状态可使用自旋轨道扭矩进行电重复控制。探究了几何形状对突触行为的影响，结果表明梯形器件具有不对称的权重更新且可控性高，而矩形器件具有更高的随机性，但电阻水平稳定。将器件数据输入神经形态计算模拟器以展示特定应用突触功能的实用性。通过实现一个应用于流数据Fashion - MNIST的人工神经网络（NN），梯形磁突触可作为一种后塑性函数用于高效在线学习。通过实现一个用于CIFAR - 100图像识别的卷积神经网络，由于其电阻水平的稳定性，矩形磁突触实现了近乎理想的推理精度。这项工作表明多权重磁突触是神经形态计算的一种可行技术，并为新兴的人工突触技术提供了设计指南。
In neuromorphic computing, artificial synapses provide a multi‐weight (MW) conductance state that is set based on inputs from neurons, analogous to the brain. Herein, artificial synapses based on magnetic materials that use a magnetic tunnel junction (MTJ) and a magnetic domain wall (DW) are explored. By fabricating lithographic notches in a DW track underneath a single MTJ, 3–5 stable resistance states that can be repeatably controlled electrically using spin‐orbit torque are achieved. The effect of geometry on the synapse behavior is explored, showing that a trapezoidal device has asymmetric weight updates with high controllability, while a rectangular device has higher stochasticity, but with stable resistance levels. The device data is input into neuromorphic computing simulators to show the usefulness of application‐specific synaptic functions. Implementing an artificial neural network (NN) applied to streamed Fashion‐MNIST data, the trapezoidal magnetic synapse can be used as a metaplastic function for efficient online learning. Implementing a convolutional NN for CIFAR‐100 image recognition, the rectangular magnetic synapse achieves near‐ideal inference accuracy, due to the stability of its resistance levels. This work shows MW magnetic synapses are a feasible technology for neuromorphic computing and provides design guidelines for emerging artificial synapse technologies.