Altmetric

Deep neural network approximation for custom hardware: where we've been, where we're going

Publication available at: https://arxiv.org/abs/1901.06955
Title: Deep neural network approximation for custom hardware: where we've been, where we're going
Authors: Wang, E
Davis, J
Zhao, R
Ng, H
Niu, X
Luk, W
Cheung, P
Constantinides, G
Item Type: Journal Article
Abstract: Deep neural networks have proven to be particularly effective in visual and audio recognition tasks. Existing models tend to be computationally expensive and memory intensive, however, and so methods for hardware-oriented approximation have become a hot topic. Research has shown that custom hardware-based neural network accelerators can surpass their general-purpose processor equivalents in terms of both throughput and energy efficiency. Application-tailored accelerators, when co-designed with approximation-based network training methods, transform large, dense and computationally expensive networks into small, sparse and hardware-efficient alternatives, increasing the feasibility of network deployment. In this article, we provide a comprehensive evaluation of approximation methods for high-performance network inference along with in-depth discussion of their effectiveness for custom hardware implementation. We also include proposals for future research based on a thorough analysis of current trends. This article represents the first survey providing detailed comparisons of custom hardware accelerators featuring approximation for both convolutional and recurrent neural networks, through which we hope to inspire exciting new developments in the field.
Issue Date: 31-May-2019
Date of Acceptance: 16-Jan-2019
URI: http://hdl.handle.net/10044/1/67297
DOI: https://dx.doi.org/10.1145/3309551
ISSN: 0360-0300
Publisher: Association for Computing Machinery
Start Page: 40:1
End Page: 40:39
Journal / Book Title: ACM Computing Surveys
Volume: 52
Issue: 2
Copyright Statement: © 2019 Association for Computing Machinery
Sponsor/Funder: Engineering & Physical Science Research Council (E
Engineering & Physical Science Research Council (EPSRC)
Engineering & Physical Science Research Council (E
Engineering & Physical Science Research Council (EPSRC)
Commission of the European Communities
Royal Academy Of Engineering
Imagination Technologies Ltd
Funder's Grant Number: 11908 (EP/K034448/1)
EP/P010040/1
516075101 (EP/N031768/1)
EP/I012036/1
671653
Prof Constantinides Chair
Prof Constantinides Chair
Keywords: Science & Technology
Technology
Computer Science, Theory & Methods
Computer Science
FPGAs
ASICs
approximation methods
convolutional neural networks
recurrent neural networks
cs.CV
cs.CV
08 Information and Computing Sciences
Information Systems
Publication Status: Published
Open Access location: https://arxiv.org/abs/1901.06955
Article Number: ARTN 40
Online Publication Date: 2019-05-31
Appears in Collections:Faculty of Engineering
Computing
Electrical and Electronic Engineering



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

Creative Commons