Sparse neuromorphic computing based on spin-torque diodes

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....

Full description

Saved in:
Bibliographic Details
Published in:Applied physics letters 2019-05, Vol.114 (19)
Main Authors: Cai, Jialin, Zhang, Like, Fang, Bin, Lv, Wenxing, Zhang, Baoshun, Finocchio, Giovanni, Xiong, Rui, Liang, Shiheng, Zeng, Zhongming
Format: Article
Language:English
Subjects:
Applied physics
Bias
Computation
Digits
Diodes
Direct current
Frequency locking
Handwriting
Handwriting recognition
Neural networks
Neuromorphic computing
Neurons
Torque
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary: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.
ISSN:0003-6951
1077-3118
DOI:10.1063/1.5090566