A 2.86-TOPS/W current mirror cross-bar-based machine-learning and physical unclonable function engine for Internet-of-Things applications

Energy-efficient machine-learning and physical unclonable function (PUF) has drawn significant attention for Internet-of-Things (IoT) application in wake-up detection for bandwidth/computation reduction and privacy protection at sensor node or autonomous device. A machine-learning and PUF engine for...

Full description

Saved in:
Bibliographic Details
Main Authors: Chen, Yi, Wang, Zheng, Patil, Aakash, Basu, Arindam
Other Authors: School of Electrical and Electronic Engineering
Format: Article
Language:English
Published: 2021
Subjects:
Online Access:https://hdl.handle.net/10356/150801
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-150801
record_format dspace
spelling sg-ntu-dr.10356-1508012021-08-02T01:36:20Z A 2.86-TOPS/W current mirror cross-bar-based machine-learning and physical unclonable function engine for Internet-of-Things applications Chen, Yi Wang, Zheng Patil, Aakash Basu, Arindam School of Electrical and Electronic Engineering VIRTUS, IC Design Centre of Excellence Engineering::Electrical and electronic engineering Machine Learning Co-processor Energy-efficient machine-learning and physical unclonable function (PUF) has drawn significant attention for Internet-of-Things (IoT) application in wake-up detection for bandwidth/computation reduction and privacy protection at sensor node or autonomous device. A machine-learning and PUF engine for IoT applications is presented in this paper with a current mirror cross-bar (CMCB) being a shared core circuit for both functions, leading to reduction in overhead area by 48.5 ×. A novel dimension expansion technique is proposed to increase weight matrix dimension beyond the physically implemented array with small hardware and energy overhead. A signed multiply-accumulation is realized in CMCB with differential current path and 2-phase conversion. The proposed engine achieves an error rate of 6.34% on MNIST digit recognition task with an energy efficiency of 2.86 TOPS/W. The PUF achieves a native bit error rate of 2.3% across corners and extremely low area per challenge response pair (CRP) of 4.17 × 10⁻⁵⁹ μm² /CRP due to exponentially more CRP enabled by ternary input mode. 2021-08-02T01:36:20Z 2021-08-02T01:36:20Z 2019 Journal Article Chen, Y., Wang, Z., Patil, A. & Basu, A. (2019). A 2.86-TOPS/W current mirror cross-bar-based machine-learning and physical unclonable function engine for Internet-of-Things applications. IEEE Transactions On Circuits and Systems I: Regular Papers, 66(6), 2240-2252. https://dx.doi.org/10.1109/TCSI.2018.2889779 1549-8328 0000-0002-4416-554X 0000-0003-2855-9570 0000-0003-1035-8770 https://hdl.handle.net/10356/150801 10.1109/TCSI.2018.2889779 2-s2.0-85065878033 6 66 2240 2252 en IEEE Transactions on Circuits and Systems I: Regular Papers © 2019 IEEE. All rights reserved.
institution Nanyang Technological University
building NTU Library
continent Asia
country Singapore
Singapore
content_provider NTU Library
collection DR-NTU
language English
topic Engineering::Electrical and electronic engineering
Machine Learning
Co-processor
spellingShingle Engineering::Electrical and electronic engineering
Machine Learning
Co-processor
Chen, Yi
Wang, Zheng
Patil, Aakash
Basu, Arindam
A 2.86-TOPS/W current mirror cross-bar-based machine-learning and physical unclonable function engine for Internet-of-Things applications
description Energy-efficient machine-learning and physical unclonable function (PUF) has drawn significant attention for Internet-of-Things (IoT) application in wake-up detection for bandwidth/computation reduction and privacy protection at sensor node or autonomous device. A machine-learning and PUF engine for IoT applications is presented in this paper with a current mirror cross-bar (CMCB) being a shared core circuit for both functions, leading to reduction in overhead area by 48.5 ×. A novel dimension expansion technique is proposed to increase weight matrix dimension beyond the physically implemented array with small hardware and energy overhead. A signed multiply-accumulation is realized in CMCB with differential current path and 2-phase conversion. The proposed engine achieves an error rate of 6.34% on MNIST digit recognition task with an energy efficiency of 2.86 TOPS/W. The PUF achieves a native bit error rate of 2.3% across corners and extremely low area per challenge response pair (CRP) of 4.17 × 10⁻⁵⁹ μm² /CRP due to exponentially more CRP enabled by ternary input mode.
author2 School of Electrical and Electronic Engineering
author_facet School of Electrical and Electronic Engineering
Chen, Yi
Wang, Zheng
Patil, Aakash
Basu, Arindam
format Article
author Chen, Yi
Wang, Zheng
Patil, Aakash
Basu, Arindam
author_sort Chen, Yi
title A 2.86-TOPS/W current mirror cross-bar-based machine-learning and physical unclonable function engine for Internet-of-Things applications
title_short A 2.86-TOPS/W current mirror cross-bar-based machine-learning and physical unclonable function engine for Internet-of-Things applications
title_full A 2.86-TOPS/W current mirror cross-bar-based machine-learning and physical unclonable function engine for Internet-of-Things applications
title_fullStr A 2.86-TOPS/W current mirror cross-bar-based machine-learning and physical unclonable function engine for Internet-of-Things applications
title_full_unstemmed A 2.86-TOPS/W current mirror cross-bar-based machine-learning and physical unclonable function engine for Internet-of-Things applications
title_sort 2.86-tops/w current mirror cross-bar-based machine-learning and physical unclonable function engine for internet-of-things applications
publishDate 2021
url https://hdl.handle.net/10356/150801
_version_ 1707050388828454912