Pattern recognition with neuromorphic computing using magnetic field-induced dynamics of skyrmions.

Pattern recognition with neuromorphic computing using magnetic field-induced dynamics of skyrmions.
复制标题

DOI:
10.1126/sciadv.abq5652
复制
发表时间:
2022-09-30
期刊:
影响因子:
13.6
通讯作者:
Otani Y
Otani Y
中科院分区:
综合性期刊1区
文献类型:
--
作者:
Yokouchi T;Sugimoto S;Rana B;Seki S;Ogawa N;Shiomi Y;Kasai S;Otani Y

文献摘要

参考文献

相似文献

物理系统中的非线性现象可以用于低能耗的脑启发计算。来自一种称为Skyrmion的拓扑自旋结构的动力学响应是这种神经形态计算的候选之一。然而,它的能力还没有得到很好的实验探索。在这里,我们实验地演示了使用源于磁场诱导的天子动力学的非线性响应的神经形态计算。我们设计了一个结构简单的基于Skyrmion的神经形态识别装置,并成功地进行了手写数字识别和波形识别,准确率高达94.7%。值得注意的是,识别准确率与设备中的Skyrmions数量之间存在正相关。天米子系统的大自由度,如位置和大小,源于更复杂的非线性映射,更大的输出维度,因此,高精度。我们的结果为开发节能、高性能的Skyrmion神经形态计算设备提供了指导。基于Skyrmion的神经形态计算设备以高精度识别波形和手写数字。
Nonlinear phenomena in physical systems can be used for brain-inspired computing with low energy consumption. Response from the dynamics of a topological spin structure called skyrmion is one of the candidates for such a neuromorphic computing. However, its ability has not been well explored experimentally. Here, we experimentally demonstrate neuromorphic computing using nonlinear response originating from magnetic field–induced dynamics of skyrmions. We designed a simple-structured skyrmion-based neuromorphic device and succeeded in handwritten digit recognition with the accuracy as large as 94.7% and waveform recognition. Notably, there exists a positive correlation between the recognition accuracy and the number of skyrmions in the devices. The large degrees of freedom of skyrmion systems, such as the position and the size, originate from the more complex nonlinear mapping, the larger output dimension, and, thus, high accuracy. Our results provide a guideline for developing energy-saving and high-performance skyrmion neuromorphic computing devices. Skyrmion-based neuromorphic computing device recognizes waveforms and handwritten digits with high accuracy.
DOI: 10.1088/1361-6528/ac3f14
发表时间: 2022-03-12
期刊: NANOTECHNOLOGY
影响因子: 3.5
作者:
Chen, Chao;Lin, Tao;Lei, Na
通讯作者: Lei, Na
DOI: 10.1038/nmat4566
发表时间: 2016-05-01
期刊: NATURE MATERIALS
影响因子: 41.2
作者:
Fukami, Shunsuke;Zhang, Chaoliang;Ohno, Hideo
通讯作者: Ohno, Hideo
DOI: 10.1126/science.aaa1442
发表时间: 2015-07-17
期刊: SCIENCE
影响因子: 56.9
作者:
Jiang, Wanjun;Upadhyaya, Pramey;Hoffmann, Axel
通讯作者: Hoffmann, Axel
使用磁性斯格明子忆阻器和自旋扭矩纳米振荡器进行物理储层计算
DOI: 10.1063/1.5115183
发表时间: 2019-11-04
影响因子: 4
作者:
Jiang, Wencong;Chen, Lina;Liu, R. H.
通讯作者: Liu, R. H.
DOI: 10.1063/1.5079305
发表时间: 2019-01-07
影响因子: 4
作者:
Markovic, D.;Leroux, N.;Grollier, J.
通讯作者: Grollier, J.