Please use this identifier to cite or link to this item:
Title: Scalable multistate spin orbit torque memory device for hopfield network
Authors: Tan, Malvin Yee Phang
Keywords: DRNTU::Science::Physics
Issue Date: 2018
Abstract: In this project, a Hopfield artificial neural network based on a non-volatile solid state multistate spintronics device is developed. The device is driven by the spin orbit torque (SOT) switching mechanism. There is a research gap on the horizontal scalability therefore the project’s objective is to investigate into the horizontal scalability of a system of SOT devices. The system consists of 10 cartridges where each cartridge holds 10 SOT devices. The cartridges are designed using computer aided design software and produced by a printed circuit board manufacturer. The device is capable of storing multistate memory where each device represents a connectivity weight of the Hopfield model. To read and write onto the device, we will be using a Keithley 2400 as the voltmeter and current source. An Arduino controller board is used to send the command of which one device out of 100 devices to access. An initial characterization of the hysteresis loop of each individual device, once done we will have a system of spintronics memory device that is capable of a Hopfield model operation.
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:SPMS Student Reports (FYP/IA/PA/PI)

Files in This Item:
File Description SizeFormat 
  Restricted Access
Report1.58 MBAdobe PDFView/Open

Page view(s) 5

checked on Oct 23, 2020

Download(s) 5

checked on Oct 23, 2020

Google ScholarTM


Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.