A Current Mirror Cross Bar Based 2.86-TOPS/W Machine Learner and PUF with <2.5% BER in 65nm CMOS for IoT Application

Yi Chen; Zheng Wang; Aakash Patil; Arindam Basu

doi:10.1109/ISCAS.2019.8702441

A Current Mirror Cross Bar Based 2.86-TOPS/W Machine Learner and PUF with <2.5% BER in 65nm CMOS for IoT Application

Yi Chen
, Zheng Wang
, Aakash Patil
, Arindam Basu

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

2 Citations (Scopus)

Abstract

Energy-efficient machine-learning and physical unclonable function (PUF) becomes popular in Internet-of-Things (IoT) applications for saliency detection and privacy protection at sensor node. A machine-learning and PUF engine for IoT applications is presented in this work with a current mirror cross-bar (CMCB) array being a shared core, reducing silicon area. 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-and-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 BER of 2.3% across corners and extremely low area/CRP of 4.17×10−⁵⁹ µm²/CRP.

Original language	English
Title of host publication	2019 IEEE International Symposium on Circuits and Systems, ISCAS 2019 - Proceedings
Publisher	IEEE
Volume	2019-May
ISBN (Print)	9781728103976
DOIs	https://doi.org/10.1109/ISCAS.2019.8702441
Publication status	Published - May 2019
Externally published	Yes
Event	2019 IEEE International Symposium on Circuits and Systems, ISCAS 2019 - Sapporo, Japan Duration: 26 May 2019 → 29 May 2019

Publication series

Name	Proceedings - IEEE International Symposium on Circuits and Systems
Volume	2019-May
ISSN (Print)	0271-4310

Conference

Conference	2019 IEEE International Symposium on Circuits and Systems, ISCAS 2019
Place	Japan
City	Sapporo
Period	26/05/19 → 29/05/19

Bibliographical note

Publication details (e.g. title, author(s), publication statuses and dates) are captured on an “AS IS” and “AS AVAILABLE” basis at the time of record harvesting from the data source. Suggestions for further amendments or supplementary information can be sent to [email protected].

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 7 Affordable and Clean Energy

Research Keywords

Crossbar
Current mirror
Edge computing
Extreme Learning Machine
Internet-Of-Things
Machine learning
Physical unclonable function

Access Output

10.1109/ISCAS.2019.8702441

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

Chen, Y., Wang, Z., Patil, A., & Basu, A. (2019). A Current Mirror Cross Bar Based 2.86-TOPS/W Machine Learner and PUF with <2.5% BER in 65nm CMOS for IoT Application. In 2019 IEEE International Symposium on Circuits and Systems, ISCAS 2019 - Proceedings (Vol. 2019-May). Article 8702441 (Proceedings - IEEE International Symposium on Circuits and Systems; Vol. 2019-May). IEEE. https://doi.org/10.1109/ISCAS.2019.8702441

@inproceedings{f8ae6780acad40d3acca24ffbb4f26c9,

title = "A Current Mirror Cross Bar Based 2.86-TOPS/W Machine Learner and PUF with <2.5\% BER in 65nm CMOS for IoT Application",

abstract = "Energy-efficient machine-learning and physical unclonable function (PUF) becomes popular in Internet-of-Things (IoT) applications for saliency detection and privacy protection at sensor node. A machine-learning and PUF engine for IoT applications is presented in this work with a current mirror cross-bar (CMCB) array being a shared core, reducing silicon area. 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-and-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 BER of 2.3\% across corners and extremely low area/CRP of 4.17×10−59 µm2/CRP.",

keywords = "Crossbar, Current mirror, Edge computing, Extreme Learning Machine, Internet-Of-Things, Machine learning, Physical unclonable function",

author = "Yi Chen and Zheng Wang and Aakash Patil and Arindam Basu",

note = "Publication details (e.g. title, author(s), publication statuses and dates) are captured on an “AS IS” and “AS AVAILABLE” basis at the time of record harvesting from the data source. Suggestions for further amendments or supplementary information can be sent to [email protected]. ; 2019 IEEE International Symposium on Circuits and Systems, ISCAS 2019 ; Conference date: 26-05-2019 Through 29-05-2019",

year = "2019",

month = may,

doi = "10.1109/ISCAS.2019.8702441",

language = "English",

isbn = "9781728103976",

volume = "2019-May",

series = "Proceedings - IEEE International Symposium on Circuits and Systems",

publisher = "IEEE",

booktitle = "2019 IEEE International Symposium on Circuits and Systems, ISCAS 2019 - Proceedings",

address = "United States",

}

Chen, Y, Wang, Z, Patil, A & Basu, A 2019, A Current Mirror Cross Bar Based 2.86-TOPS/W Machine Learner and PUF with <2.5% BER in 65nm CMOS for IoT Application. in 2019 IEEE International Symposium on Circuits and Systems, ISCAS 2019 - Proceedings. vol. 2019-May, 8702441, Proceedings - IEEE International Symposium on Circuits and Systems, vol. 2019-May, IEEE, 2019 IEEE International Symposium on Circuits and Systems, ISCAS 2019, Sapporo, Japan, 26/05/19. https://doi.org/10.1109/ISCAS.2019.8702441

A Current Mirror Cross Bar Based 2.86-TOPS/W Machine Learner and PUF with <2.5% BER in 65nm CMOS for IoT Application. / Chen, Yi; Wang, Zheng; Patil, Aakash et al.
2019 IEEE International Symposium on Circuits and Systems, ISCAS 2019 - Proceedings. Vol. 2019-May IEEE, 2019. 8702441 (Proceedings - IEEE International Symposium on Circuits and Systems; Vol. 2019-May).

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

TY - GEN

T1 - A Current Mirror Cross Bar Based 2.86-TOPS/W Machine Learner and PUF with <2.5% BER in 65nm CMOS for IoT Application

AU - Chen, Yi

AU - Wang, Zheng

AU - Patil, Aakash

AU - Basu, Arindam

N1 - Publication details (e.g. title, author(s), publication statuses and dates) are captured on an “AS IS” and “AS AVAILABLE” basis at the time of record harvesting from the data source. Suggestions for further amendments or supplementary information can be sent to [email protected].

PY - 2019/5

Y1 - 2019/5

N2 - Energy-efficient machine-learning and physical unclonable function (PUF) becomes popular in Internet-of-Things (IoT) applications for saliency detection and privacy protection at sensor node. A machine-learning and PUF engine for IoT applications is presented in this work with a current mirror cross-bar (CMCB) array being a shared core, reducing silicon area. 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-and-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 BER of 2.3% across corners and extremely low area/CRP of 4.17×10−59 µm2/CRP.

AB - Energy-efficient machine-learning and physical unclonable function (PUF) becomes popular in Internet-of-Things (IoT) applications for saliency detection and privacy protection at sensor node. A machine-learning and PUF engine for IoT applications is presented in this work with a current mirror cross-bar (CMCB) array being a shared core, reducing silicon area. 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-and-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 BER of 2.3% across corners and extremely low area/CRP of 4.17×10−59 µm2/CRP.

KW - Crossbar

KW - Current mirror

KW - Edge computing

KW - Extreme Learning Machine

KW - Internet-Of-Things

KW - Machine learning

KW - Physical unclonable function

UR - https://www.scopus.com/pages/publications/85066810177

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-85066810177&origin=recordpage

U2 - 10.1109/ISCAS.2019.8702441

DO - 10.1109/ISCAS.2019.8702441

M3 - RGC 32 - Refereed conference paper (with host publication)

SN - 9781728103976

VL - 2019-May

T3 - Proceedings - IEEE International Symposium on Circuits and Systems

BT - 2019 IEEE International Symposium on Circuits and Systems, ISCAS 2019 - Proceedings

PB - IEEE

T2 - 2019 IEEE International Symposium on Circuits and Systems, ISCAS 2019

Y2 - 26 May 2019 through 29 May 2019

ER -

Chen Y, Wang Z, Patil A, Basu A. A Current Mirror Cross Bar Based 2.86-TOPS/W Machine Learner and PUF with <2.5% BER in 65nm CMOS for IoT Application. In 2019 IEEE International Symposium on Circuits and Systems, ISCAS 2019 - Proceedings. Vol. 2019-May. IEEE. 2019. 8702441. (Proceedings - IEEE International Symposium on Circuits and Systems). doi: 10.1109/ISCAS.2019.8702441

A Current Mirror Cross Bar Based 2.86-TOPS/W Machine Learner and PUF with <2.5% BER in 65nm CMOS for IoT Application

Abstract

Publication series

Conference

Bibliographical note

UN SDGs

Research Keywords

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this