DeLight: Adding Energy Dimension To Deep Neural Networks

B. Rouhani, Azalia Mirhoseini, F. Koushanfar
{"title":"DeLight: Adding Energy Dimension To Deep Neural Networks","authors":"B. Rouhani, Azalia Mirhoseini, F. Koushanfar","doi":"10.1145/2934583.2934599","DOIUrl":null,"url":null,"abstract":"Physical viability, in particular energy efficiency, is a key challenge in realizing the true potential of Deep Neural Networks (DNNs). In this paper, we aim to incorporate the energy dimension as a design parameter in the higher-level hierarchy of DNN training and execution to optimize for the energy resources and constraints. We use energy characterization to bound the network size in accordance to the pertinent physical resources. An automated customization methodology is proposed to adaptively conform the DNN configurations to the underlying hardware characteristics while minimally affecting the inference accuracy. The key to our approach is a new context and resource aware projection of data to a lower-dimensional embedding by which learning the correlation between data samples requires significantly smaller number of neurons. We leverage the performance gain achieved as a result of the data projection to enable the training of different DNN architectures which can be aggregated together to further boost the inference accuracy. Accompanying APIs are provided to facilitate rapid prototyping of an arbitrary DNN application customized to the underlying platform. Proof-of-concept evaluations for deployment of different visual, audio, and smart-sensing benchmarks demonstrate up to 100-fold energy improvement compared to the prior-art DL solutions.","PeriodicalId":142716,"journal":{"name":"Proceedings of the 2016 International Symposium on Low Power Electronics and Design","volume":"112 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2016-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"47","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the 2016 International Symposium on Low Power Electronics and Design","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/2934583.2934599","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 47

Abstract

Physical viability, in particular energy efficiency, is a key challenge in realizing the true potential of Deep Neural Networks (DNNs). In this paper, we aim to incorporate the energy dimension as a design parameter in the higher-level hierarchy of DNN training and execution to optimize for the energy resources and constraints. We use energy characterization to bound the network size in accordance to the pertinent physical resources. An automated customization methodology is proposed to adaptively conform the DNN configurations to the underlying hardware characteristics while minimally affecting the inference accuracy. The key to our approach is a new context and resource aware projection of data to a lower-dimensional embedding by which learning the correlation between data samples requires significantly smaller number of neurons. We leverage the performance gain achieved as a result of the data projection to enable the training of different DNN architectures which can be aggregated together to further boost the inference accuracy. Accompanying APIs are provided to facilitate rapid prototyping of an arbitrary DNN application customized to the underlying platform. Proof-of-concept evaluations for deployment of different visual, audio, and smart-sensing benchmarks demonstrate up to 100-fold energy improvement compared to the prior-art DL solutions.
喜悦:为深度神经网络添加能量维度
物理可行性,特别是能源效率,是实现深度神经网络(dnn)真正潜力的关键挑战。在本文中,我们的目标是将能量维度作为设计参数纳入深度神经网络训练和执行的更高层次结构中,以优化能量资源和约束。我们使用能量表征来根据相关的物理资源约束网络大小。提出了一种自动自定义方法,使深度神经网络配置自适应地符合底层硬件特征,同时对推理精度的影响最小。我们方法的关键是一种新的上下文和资源感知的数据投影到低维嵌入,通过这种嵌入学习数据样本之间的相关性需要更少的神经元数量。我们利用数据投影所获得的性能增益来训练不同的深度神经网络架构,这些架构可以聚合在一起,以进一步提高推理精度。提供了随附的api,以方便根据底层平台定制任意DNN应用程序的快速原型。针对不同视觉、音频和智能传感基准部署的概念验证评估表明,与现有技术的深度学习解决方案相比,其能耗提高了100倍。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信