Affiliation:
1. Imperial College London, London, United Kingdom
2. Corerain Technologies, Shenzhen, China
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.
Funder
China Scholarship Council
Intel Corporation
Lee Foundation
Royal Academy of Engineering
Horizon 2020
Engineering and Physical Sciences Research Council
Publisher
Association for Computing Machinery (ACM)
Subject
General Computer Science,Theoretical Computer Science
Reference163 articles.
1. Intel AI. 2017. Intel Nervana Neural Network Processors (NNP) Redefine AI Silicon. Retrieved from https://ai.intel.com/intel-nervana-neural-network-processors-nnp-redefine-ai-silicon/. Intel AI. 2017. Intel Nervana Neural Network Processors (NNP) Redefine AI Silicon. Retrieved from https://ai.intel.com/intel-nervana-neural-network-processors-nnp-redefine-ai-silicon/.
2. TrueNorth: Design and tool flow of a 65mW 1 million neuron programmable neurosynaptic chip;Akopyan Filipp;IEEE Trans. Comput.-aided Design Integr. Circ. Syst.,2015
3. Bit-pragmatic deep neural network computing
4. Cnvlutin: Ineffectual-Neuron-Free Deep Neural Network Computing
5. Ternary neural networks for resource-efficient AI applications
Cited by
112 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献