InstantNet: Automated Generation and Deployment of Instantaneously Switchable-Precision Networks

Yonggan Fu; Zhongzhi Yu; Yongan Zhang; Yifan Jiang; Chaojian Li; Yongyuan Liang; Mingchao Jiang; Zhangyang Wang; Yingyan Lin

doi:10.1109/DAC18074.2021.9586254

InstantNet: Automated Generation and Deployment of Instantaneously Switchable-Precision Networks

Yonggan Fu, Zhongzhi Yu, Yongan Zhang, Yifan Jiang, Chaojian Li, Yongyuan Liang, Mingchao Jiang, Zhangyang Wang, Yingyan Lin

Research output: Chapter in Book/Conference Proceeding/Report › Conference Paper published in a book › peer-review

2 Citations (Scopus)

Abstract

The promise of Deep Neural Network (DNN) powered Internet of Thing (IoT) devices has motivated a tremendous demand for automated solutions to enable fast development and deployment of efficient (1) DNNs equipped with instantaneous accuracy-efficiency trade-off capability to accommodate the time-varying resources at IoT devices and (2) dataflows to optimize DNNs' execution efficiency on different devices. Therefore, we propose InstantNet to automatically generate and deploy instantaneously switchable-precision networks which operates at variable bit-widths. Extensive experiments show that the proposed InstantNet consistently outperforms state-of-the-art designs. Our codes are available at: https://github.com/RICE-EIC/InstantNet.

Original language	English
Title of host publication	2021 58th ACM/IEEE Design Automation Conference, DAC 2021
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	757-762
Number of pages	6
ISBN (Electronic)	9781665432740
DOIs	https://doi.org/10.1109/DAC18074.2021.9586254
Publication status	Published - 5 Dec 2021
Externally published	Yes
Event	58th ACM/IEEE Design Automation Conference, DAC 2021 - San Francisco, United States Duration: 5 Dec 2021 → 9 Dec 2021

Publication series

Name	Proceedings - Design Automation Conference
Volume	2021-December
ISSN (Print)	0738-100X

Conference

Conference	58th ACM/IEEE Design Automation Conference, DAC 2021
Country/Territory	United States
City	San Francisco
Period	5/12/21 → 9/12/21

Bibliographical note

Publisher Copyright:
© 2021 IEEE.

Keywords

NAS
dataflow
switchable-precision networks

Access to Document

10.1109/DAC18074.2021.9586254

Cite this

Fu, Y., Yu, Z., Zhang, Y., Jiang, Y., Li, C., Liang, Y., Jiang, M., Wang, Z., & Lin, Y. (2021). InstantNet: Automated Generation and Deployment of Instantaneously Switchable-Precision Networks. In 2021 58th ACM/IEEE Design Automation Conference, DAC 2021 (pp. 757-762). (Proceedings - Design Automation Conference; Vol. 2021-December). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/DAC18074.2021.9586254

@inproceedings{0cb4ef420a7447fa893093e8f25c1863,

title = "InstantNet: Automated Generation and Deployment of Instantaneously Switchable-Precision Networks",

abstract = "The promise of Deep Neural Network (DNN) powered Internet of Thing (IoT) devices has motivated a tremendous demand for automated solutions to enable fast development and deployment of efficient (1) DNNs equipped with instantaneous accuracy-efficiency trade-off capability to accommodate the time-varying resources at IoT devices and (2) dataflows to optimize DNNs' execution efficiency on different devices. Therefore, we propose InstantNet to automatically generate and deploy instantaneously switchable-precision networks which operates at variable bit-widths. Extensive experiments show that the proposed InstantNet consistently outperforms state-of-the-art designs. Our codes are available at: https://github.com/RICE-EIC/InstantNet.",

keywords = "NAS, dataflow, switchable-precision networks",

author = "Yonggan Fu and Zhongzhi Yu and Yongan Zhang and Yifan Jiang and Chaojian Li and Yongyuan Liang and Mingchao Jiang and Zhangyang Wang and Yingyan Lin",

note = "Publisher Copyright: {\textcopyright} 2021 IEEE.; 58th ACM/IEEE Design Automation Conference, DAC 2021 ; Conference date: 05-12-2021 Through 09-12-2021",

year = "2021",

month = dec,

day = "5",

doi = "10.1109/DAC18074.2021.9586254",

language = "English",

series = "Proceedings - Design Automation Conference",

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

pages = "757--762",

booktitle = "2021 58th ACM/IEEE Design Automation Conference, DAC 2021",

}

Fu, Y, Yu, Z, Zhang, Y, Jiang, Y, Li, C, Liang, Y, Jiang, M, Wang, Z & Lin, Y 2021, InstantNet: Automated Generation and Deployment of Instantaneously Switchable-Precision Networks. in 2021 58th ACM/IEEE Design Automation Conference, DAC 2021. Proceedings - Design Automation Conference, vol. 2021-December, Institute of Electrical and Electronics Engineers Inc., pp. 757-762, 58th ACM/IEEE Design Automation Conference, DAC 2021, San Francisco, United States, 5/12/21. https://doi.org/10.1109/DAC18074.2021.9586254

InstantNet: Automated Generation and Deployment of Instantaneously Switchable-Precision Networks. / Fu, Yonggan; Yu, Zhongzhi; Zhang, Yongan et al.
2021 58th ACM/IEEE Design Automation Conference, DAC 2021. Institute of Electrical and Electronics Engineers Inc., 2021. p. 757-762 (Proceedings - Design Automation Conference; Vol. 2021-December).

Research output: Chapter in Book/Conference Proceeding/Report › Conference Paper published in a book › peer-review

TY - GEN

T1 - InstantNet

T2 - 58th ACM/IEEE Design Automation Conference, DAC 2021

AU - Fu, Yonggan

AU - Yu, Zhongzhi

AU - Zhang, Yongan

AU - Jiang, Yifan

AU - Li, Chaojian

AU - Liang, Yongyuan

AU - Jiang, Mingchao

AU - Wang, Zhangyang

AU - Lin, Yingyan

PY - 2021/12/5

Y1 - 2021/12/5

N2 - The promise of Deep Neural Network (DNN) powered Internet of Thing (IoT) devices has motivated a tremendous demand for automated solutions to enable fast development and deployment of efficient (1) DNNs equipped with instantaneous accuracy-efficiency trade-off capability to accommodate the time-varying resources at IoT devices and (2) dataflows to optimize DNNs' execution efficiency on different devices. Therefore, we propose InstantNet to automatically generate and deploy instantaneously switchable-precision networks which operates at variable bit-widths. Extensive experiments show that the proposed InstantNet consistently outperforms state-of-the-art designs. Our codes are available at: https://github.com/RICE-EIC/InstantNet.

AB - The promise of Deep Neural Network (DNN) powered Internet of Thing (IoT) devices has motivated a tremendous demand for automated solutions to enable fast development and deployment of efficient (1) DNNs equipped with instantaneous accuracy-efficiency trade-off capability to accommodate the time-varying resources at IoT devices and (2) dataflows to optimize DNNs' execution efficiency on different devices. Therefore, we propose InstantNet to automatically generate and deploy instantaneously switchable-precision networks which operates at variable bit-widths. Extensive experiments show that the proposed InstantNet consistently outperforms state-of-the-art designs. Our codes are available at: https://github.com/RICE-EIC/InstantNet.

KW - NAS

KW - dataflow

KW - switchable-precision networks

UR - https://www.scopus.com/pages/publications/85119415583

U2 - 10.1109/DAC18074.2021.9586254

DO - 10.1109/DAC18074.2021.9586254

M3 - Conference Paper published in a book

AN - SCOPUS:85119415583

T3 - Proceedings - Design Automation Conference

SP - 757

EP - 762

BT - 2021 58th ACM/IEEE Design Automation Conference, DAC 2021

PB - Institute of Electrical and Electronics Engineers Inc.

Y2 - 5 December 2021 through 9 December 2021

ER -

InstantNet: Automated Generation and Deployment of Instantaneously Switchable-Precision Networks

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this