2020
DOI: 10.1109/access.2020.3032188
|View full text |Cite
|
Sign up to set email alerts
|

Digital and Analog Switching Characteristics of InGaZnO Memristor Depending on Top Electrode Material for Neuromorphic System

Abstract: In this study, we demonstrate both of digital and analog memory operations in InGaZnO (IGZO) memristor devices by controlling the electrode materials for neuromorphic application. The switching properties of the devices are determined by the initial energy barrier characteristics between the metal electrodes and the IGZO switching layer. Digital switching characteristics are obtained after the forming process when Schottky junction occurs at both of top and bottom electrodes. On the other hands, analog resisti… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
2

Citation Types

0
18
0

Year Published

2021
2021
2024
2024

Publication Types

Select...
9

Relationship

2
7

Authors

Journals

citations
Cited by 31 publications
(21 citation statements)
references
References 49 publications
0
18
0
Order By: Relevance
“…Here, the network is trained with the gradient descent algorithm assuming on-chip learning with the measured learning characteristics in Figure b,c. How large each weight should be updated (Δ W ) is determined by the weight-update model developed with the relation between the amount of weight change and current weight value . First, the nonlinear weight modulation with the pulse condition A significantly degrades the recognition rate (dotted lines) no matter which kind of device is used as a synaptic transistor.…”
Section: Results and Discussionmentioning
confidence: 99%
See 1 more Smart Citation
“…Here, the network is trained with the gradient descent algorithm assuming on-chip learning with the measured learning characteristics in Figure b,c. How large each weight should be updated (Δ W ) is determined by the weight-update model developed with the relation between the amount of weight change and current weight value . First, the nonlinear weight modulation with the pulse condition A significantly degrades the recognition rate (dotted lines) no matter which kind of device is used as a synaptic transistor.…”
Section: Results and Discussionmentioning
confidence: 99%
“…How large each weight should be updated (ΔW) is determined by the weight-update model developed with the relation between the amount of weight change and current weight value. 52 First, the nonlinear weight modulation with the pulse condition A significantly degrades the recognition rate (dotted lines) no matter which kind of device is used as a synaptic transistor. Especially, the accuracy is not saturated at all after enough number of epochs with the pulse condition A.…”
Section: ■ Results and Discussionmentioning
confidence: 99%
“…Oxidation and reduction reactions occur inside IGZO depending on the external bias, which changes the resistance of IGZO; thus, IGZO is used as a memory device by utilizing these characteristics [ 42 , 43 ]. In fact, there have already been studies on various memory devices using IGZO [ 44 , 45 , 46 , 47 , 48 , 49 ]. In particular, the p + Si-based IGZO memristors introduced in this paper show high retention and endurance characteristics.…”
Section: Introductionmentioning
confidence: 99%
“…Memristor can be a building block in neuromorphic systems by controlling conductance [ 7 , 8 , 9 , 10 , 11 , 12 , 13 ]. Various resistive switching behaviors are observed in a lot of materials, such as oxides, nitrides and other materials [ 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 ]. Among them, the indium gallium zinc oxide (IGZO)-based memristor is very promising for memory applications because of its interface-type switching and good compatibility with conventional Si processing [ 14 , 15 , 16 , 17 , 18 , 19 ].…”
Section: Introductionmentioning
confidence: 99%