Sparse neuromorphic computing based on spin-torque diodes

被引：37

作者：

Cai, Jialin ^{[1
,2
]}

Zhang, Like ^{[1
,2
]}

Fang, Bin ^{[1
]}

Lv, Wenxing ^{[1
,2
]}

Zhang, Baoshun ^{[1
,2
]}

Finocchio, Giovanni ^{[3
]}

Xiong, Rui ^{[4
]}

Liang, Shiheng ^{[5
]}

Zeng, Zhongming ^{[1
,2
]}

机构：

[1] Chinese Acad Sci, Suzhou Inst Nanotech & Nanobion, Key Lab Multifunct Nanomat & Smart Syst, Suzhou 215123, Jiangsu, Peoples R China

[2] Univ Sci & Technol China, Sch Nano Technol & Nano Bion, Hefei 230026, Anhui, Peoples R China

[3] Univ Messina, Dept Math & Comp Sci Phys Sci & Earth Sci, I-98166 Messina, Italy

[4] Wuhan Univ, Sch Phys & Technol, Minist Educ, Key Lab Artificial Micro & Nanostruct, Wuhan 430072, Hubei, Peoples R China

[5] Hubei Univ, Dept Phys, Wuhan 430062, Hubei, Peoples R China

来源：

APPLIED PHYSICS LETTERS | 2019年 / 114卷 / 19期

基金：

美国国家科学基金会; 中国博士后科学基金;

关键词：

D O I：

10.1063/1.5090566

中图分类号：

O59 [应用物理学];

学科分类号：

摘要：

We report on the sparse neuromorphic computing based on spin-torque diodes (STDs). The rectification characteristics of STDs have been investigated in the absence and presence of d.c. bias currents. While the injection locking phenomenon is observed in our devices, the output functions versus the d.c. bias currents mimic artificial neurons with sparse representations. Furthermore, we construct a neural network with STD neurons to recognize the handwritten digits in the Mixed National Institute of Standards and Technology database, with a produced accuracy of up to 92.7%. The results suggest that STDs have potential to be building blocks for the realization of a biologically plausible neuromorphic computing system.

引用

页数：4

共 35 条

[1]

[Anonymous], 2010, INT C MACH LEARN

[2] An energy budget for signaling in the grey matter of the brain [J].