Gradient Descent on Multilevel Spin-Orbit Synapses with Tunable Variations
Author(s): Lan, XK (Lan, Xiukai); Cao, Y (Cao, Yi); Liu, XY (Liu, Xiangyu); Xu, KJ (Xu, Kaijia); Liu, C (Liu, Chuan); Zheng, HZ (Zheng, Houzhi); Wang, KY (Wang, Kaiyou)
Source: ADVANCED INTELLIGENT SYSTEMS Volume: 3 Issue: 6 Article Number: 2000182 DOI: 10.1002/aisy.202000182 Published: JUN 2021
Abstract: Neuromorphic computing using multilevel nonvolatile memories as synapses offers opportunities for future energy- and area-efficient artificial intelligence. Among these memories, artificial synapses based on current-induced magnetization switching driven by spin-orbit torques (SOTs) have attracted great attention recently. Herein, the gradient descent algorithm, a primary learning algorithm, implemented on a 2x1 SOT synaptic array is reported. Successful pattern classifications are experimentally realized through the tuning of cycle-to-cycle variation, linearity range, and linearity deviation of the multilevel SOT synapse. Also, a larger mxn SOT synaptic array with m controlling transistors is proposed and it is found that the classification accuracies can be improved dramatically by decreasing the cycle-to-cycle variation. A way for the application of spin-orbit device arrays in neuromorphic computing is paved and the crucial importance of the cycle-to-cycle variation for a multilevel SOT synapse is suggested.
Accession Number: WOS:000669819100002
Author Identifiers:
Author Web of Science ResearcherID ORCID Number
Wang, Kaiyou V-3573-2019
Lan, XiuKai 0000-0002-0068-0603
eISSN: 2640-4567
Full Text: https://onlinelibrary.wiley.com/doi/10.1002/aisy.202000182