Sub-nA Low-Current HZO Ferroelectric Tunnel Junction for High-Performance and Accurate Deep Learning Acceleration

Tzu Yun Wu; Tian-Sheuan Chang; Heng Yuan Lee; Shyh Shyuan Sheu; Wei Chung Lo; Tuo-Hung Hou; Hsin Hui Huang; Yueh Hua Chu; Chih Cheng Chang; Ming Hung Wu; Chien Hua Hsu; Chien Ting Wu; Min Ci Wu; Wen-Wei Wu

doi:10.1109/IEDM19573.2019.8993565

Sub-nA Low-Current HZO Ferroelectric Tunnel Junction for High-Performance and Accurate Deep Learning Acceleration

Tzu Yun Wu, Tian-Sheuan Chang, Heng Yuan Lee, Shyh Shyuan Sheu, Wei Chung Lo, Tuo-Hung Hou^*, Hsin Hui Huang, Yueh Hua Chu, Chih Cheng Chang, Ming Hung Wu, Chien Hua Hsu, Chien Ting Wu, Min Ci Wu, Wen-Wei Wu

^*此作品的通信作者

研究成果: Conference contribution › 同行評審

18 引文斯高帕斯（Scopus）

摘要

This paper presents a unique opportunity of HZO ferroelectric tunnel junction (FTJ) for in-memory computing. The device operates at an extremely low sub-nA current while simultaneously achieving 50-ns fast switching, > 10⁷ cycling endurance, > 10-yr retention, minimal variability, and analog state modulation. We analyze an FTJ-based deep binary neural network. It achieves better accuracy and remarkable 702, 101, and 7×10⁴ times improvements in power, area, and energy-area product efficiency compared with those using NVMs with a typical μA cell current designed for fast memory access.

原文	English
主出版物標題	2019 IEEE International Electron Devices Meeting, IEDM 2019
發行者	Institute of Electrical and Electronics Engineers Inc.
頁數	4
ISBN（電子）	9781728140315
DOIs	https://doi.org/10.1109/IEDM19573.2019.8993565
出版狀態	Published - 12月 2019
事件	65th Annual IEEE International Electron Devices Meeting, IEDM 2019 - San Francisco, United States 持續時間: 7 12月 2019 → 11 12月 2019

出版系列

名字	Technical Digest - International Electron Devices Meeting, IEDM
卷	2019-December
ISSN（列印）	0163-1918

Conference

Conference	65th Annual IEEE International Electron Devices Meeting, IEDM 2019
國家/地區	United States
城市	San Francisco
期間	7/12/19 → 11/12/19

存取文件

10.1109/IEDM19573.2019.8993565

其它文件與鏈接

Link to publication in Scopus

引用此

Wu, T. Y., Chang, T-S., Lee, H. Y., Sheu, S. S., Lo, W. C., Hou, T-H., Huang, H. H., Chu, Y. H., Chang, C. C., Wu, M. H., Hsu, C. H., Wu, C. T., Wu, M. C., & Wu, W-W. (2019). Sub-nA Low-Current HZO Ferroelectric Tunnel Junction for High-Performance and Accurate Deep Learning Acceleration. 於 2019 IEEE International Electron Devices Meeting, IEDM 2019 [8993565] (Technical Digest - International Electron Devices Meeting, IEDM; 卷 2019-December). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/IEDM19573.2019.8993565

@inproceedings{6da43cd130d6408db141f555752f915b,

title = "Sub-nA Low-Current HZO Ferroelectric Tunnel Junction for High-Performance and Accurate Deep Learning Acceleration",

abstract = "This paper presents a unique opportunity of HZO ferroelectric tunnel junction (FTJ) for in-memory computing. The device operates at an extremely low sub-nA current while simultaneously achieving 50-ns fast switching, > 107 cycling endurance, > 10-yr retention, minimal variability, and analog state modulation. We analyze an FTJ-based deep binary neural network. It achieves better accuracy and remarkable 702, 101, and 7×104 times improvements in power, area, and energy-area product efficiency compared with those using NVMs with a typical μA cell current designed for fast memory access.",

author = "Wu, {Tzu Yun} and Tian-Sheuan Chang and Lee, {Heng Yuan} and Sheu, {Shyh Shyuan} and Lo, {Wei Chung} and Tuo-Hung Hou and Huang, {Hsin Hui} and Chu, {Yueh Hua} and Chang, {Chih Cheng} and Wu, {Ming Hung} and Hsu, {Chien Hua} and Wu, {Chien Ting} and Wu, {Min Ci} and Wen-Wei Wu",

year = "2019",

month = dec,

doi = "10.1109/IEDM19573.2019.8993565",

language = "English",

series = "Technical Digest - International Electron Devices Meeting, IEDM",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "2019 IEEE International Electron Devices Meeting, IEDM 2019",

address = "United States",

note = "null ; Conference date: 07-12-2019 Through 11-12-2019",

}

Wu, TY, Chang, T-S, Lee, HY, Sheu, SS, Lo, WC, Hou, T-H, Huang, HH, Chu, YH, Chang, CC, Wu, MH, Hsu, CH, Wu, CT, Wu, MC & Wu, W-W 2019, Sub-nA Low-Current HZO Ferroelectric Tunnel Junction for High-Performance and Accurate Deep Learning Acceleration. 於 2019 IEEE International Electron Devices Meeting, IEDM 2019., 8993565, Technical Digest - International Electron Devices Meeting, IEDM, 卷 2019-December, Institute of Electrical and Electronics Engineers Inc., 65th Annual IEEE International Electron Devices Meeting, IEDM 2019, San Francisco, United States, 7/12/19. https://doi.org/10.1109/IEDM19573.2019.8993565

Sub-nA Low-Current HZO Ferroelectric Tunnel Junction for High-Performance and Accurate Deep Learning Acceleration. / Wu, Tzu Yun; Chang, Tian-Sheuan; Lee, Heng Yuan 等.

2019 IEEE International Electron Devices Meeting, IEDM 2019. Institute of Electrical and Electronics Engineers Inc., 2019. 8993565 (Technical Digest - International Electron Devices Meeting, IEDM; 卷 2019-December).

研究成果: Conference contribution › 同行評審

TY - GEN

T1 - Sub-nA Low-Current HZO Ferroelectric Tunnel Junction for High-Performance and Accurate Deep Learning Acceleration

AU - Wu, Tzu Yun

AU - Chang, Tian-Sheuan

AU - Lee, Heng Yuan

AU - Sheu, Shyh Shyuan

AU - Lo, Wei Chung

AU - Hou, Tuo-Hung

AU - Huang, Hsin Hui

AU - Chu, Yueh Hua

AU - Chang, Chih Cheng

AU - Wu, Ming Hung

AU - Hsu, Chien Hua

AU - Wu, Chien Ting

AU - Wu, Min Ci

AU - Wu, Wen-Wei

PY - 2019/12

Y1 - 2019/12

N2 - This paper presents a unique opportunity of HZO ferroelectric tunnel junction (FTJ) for in-memory computing. The device operates at an extremely low sub-nA current while simultaneously achieving 50-ns fast switching, > 107 cycling endurance, > 10-yr retention, minimal variability, and analog state modulation. We analyze an FTJ-based deep binary neural network. It achieves better accuracy and remarkable 702, 101, and 7×104 times improvements in power, area, and energy-area product efficiency compared with those using NVMs with a typical μA cell current designed for fast memory access.

AB - This paper presents a unique opportunity of HZO ferroelectric tunnel junction (FTJ) for in-memory computing. The device operates at an extremely low sub-nA current while simultaneously achieving 50-ns fast switching, > 107 cycling endurance, > 10-yr retention, minimal variability, and analog state modulation. We analyze an FTJ-based deep binary neural network. It achieves better accuracy and remarkable 702, 101, and 7×104 times improvements in power, area, and energy-area product efficiency compared with those using NVMs with a typical μA cell current designed for fast memory access.

UR - http://www.scopus.com/inward/record.url?scp=85081044261&partnerID=8YFLogxK

U2 - 10.1109/IEDM19573.2019.8993565

DO - 10.1109/IEDM19573.2019.8993565

M3 - Conference contribution

AN - SCOPUS:85081044261

T3 - Technical Digest - International Electron Devices Meeting, IEDM

BT - 2019 IEEE International Electron Devices Meeting, IEDM 2019

PB - Institute of Electrical and Electronics Engineers Inc.

Y2 - 7 December 2019 through 11 December 2019

ER -

Wu TY, Chang T-S, Lee HY, Sheu SS, Lo WC, Hou T-H 等. Sub-nA Low-Current HZO Ferroelectric Tunnel Junction for High-Performance and Accurate Deep Learning Acceleration. 於 2019 IEEE International Electron Devices Meeting, IEDM 2019. Institute of Electrical and Electronics Engineers Inc. 2019. 8993565. (Technical Digest - International Electron Devices Meeting, IEDM). doi: 10.1109/IEDM19573.2019.8993565

Sub-nA Low-Current HZO Ferroelectric Tunnel Junction for High-Performance and Accurate Deep Learning Acceleration

摘要

出版系列

Conference

存取文件

其它文件與鏈接

指紋

引用此