A study of deep learning model deployment methods at the edge

Xuejun Jin; Changbao Xu; Mingyong Xin

doi:10.1117/12.3011496

7 December 2023 A study of deep learning model deployment methods at the edge

Xuejun Jin, Changbao Xu, Mingyong Xin

Proceedings Volume 12941, International Conference on Algorithms, High Performance Computing, and Artificial Intelligence (AHPCAI 2023); 129414I (2023) https://doi.org/10.1117/12.3011496
Event: Third International Conference on Algorithms, High Performance Computing, and Artificial Intelligence (AHPCAI 203), 2023, Yinchuan, China

Abstract

Driven by the development of power IoT and the digitization of energy, the scope of application of power edge intelligent terminals is expanding. Domestic AI chips are still inadequate in covering the reasoning of various AI algorithms at the edge end of power IoT, and the support of deep learning framework for domestic chips is limited, leading to difficulties in obtaining underlying hardware acceleration. In order to achieve the goal of effective deployment of complex deep learning models in power operation safety control scenarios to domestic chips, this paper summarizes the methods of multi-edge chip co-computation, computational graph optimization, and model quantification to realize the configuration of models on edge terminals.

(2023) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Xuejun Jin, Changbao Xu, and Mingyong Xin "A study of deep learning model deployment methods at the edge", Proc. SPIE 12941, International Conference on Algorithms, High Performance Computing, and Artificial Intelligence (AHPCAI 2023), 129414I (7 December 2023); https://doi.org/10.1117/12.3011496

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available