Super-steep synapses based on positive feedback devices for reliable binary neural networks-Reference-Cited by-同舟云学术

Super-steep synapses based on positive feedback devices for reliable binary neural networks

Published:2023-03-06 Issue:10 Volume:122 Page:
ISSN:0003-6951
Container-title:Applied Physics Letters
language:en
Short-container-title:

Author:

Kwon Dongseok¹^ORCID,Kim Hyeongsu¹^ORCID,Lee Kyu-Ho¹^ORCID,Hwang Joon¹^ORCID,Shin Wonjun¹^ORCID,Bae Jong-Ho²^ORCID,Woo Sung Yun³^ORCID,Lee Jong-Ho⁴^ORCID

Affiliation:

1. Inter-University Semiconductor Research Center, Department of Electrical and Computer Engineering, Seoul National University 1 , Seoul 08826, South Korea

2. School of Electrical Engineering, Kookmin University 2 , Seoul 02707, Korea

3. School of Electronics Engineering, Kyungpook National University 3 , 80 Daehakro, Bukgu, Daegu 41566, South Korea

4. Ministry of Science and ICT 4 , Sejong, Korea

Abstract

This work proposes positive feedback (PF) device-based synaptic devices for reliable binary neural networks (BNNs). Due to PF operation, the fabricated PF device shows a high on/off current ratio (2.69 × 107). The PF device has a charge-trap layer by which the turn-on voltage (Von) of the device can be adjusted by program/erase operations and a long-term memory function is implemented. Also, due to the steep switching characteristics of the PF device, the conductance becomes tolerant to the retention time and the variation in turn-on voltage. Simulations show that high accuracy (88.44% for CIFAR-10 image classification) can be achieved in hardware-based BNNs using PF devices with these properties as synapses.

Funder

Institute of Information and Communications Technology Planning and Evaluation

Publisher

AIP Publishing

Subject

Physics and Astronomy (miscellaneous)

Link

https://pubs.aip.org/aip/apl/article-pdf/doi/10.1063/5.0131235/16773356/102101_1_online.pdf

Reference33 articles.

1. Accelerating Deep Neural Network In-Situ Training With Non-Volatile and Volatile Memory Based Hybrid Precision Synapses

2. Efficient fusion of spiking neural networks and FET-type gas sensors for a fast and reliable artificial olfactory system

3. Review of Analog Neuron Devices for Hardware-based Spiking Neural Networks

4. Multimodal Machine Learning: A Survey and Taxonomy

5. Equivalent-accuracy accelerated neural-network training using analogue memory

Cited by 2 articles. 订阅此论文施引文献订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献

1. Vertical AND-Type Flash Synaptic Cell Stack for High-Density and Reliable Binary Neural Networks;IEEE Electron Device Letters;2024-07

2. Hardware-Based Ternary Neural Network Using AND-Type Poly-Si TFT Array and Its Optimization Guideline;IEEE Transactions on Electron Devices;2023-08