PuDianNao: A Polyvalent Machine Learning Accelerator

doi:10.1145/2694344.2694358

CORC > 计算技术研究所 > 中国科学院计算技术研究所 > 中国科学院计算技术研究所期刊论文 > 英文

	PuDianNao: A Polyvalent Machine Learning Accelerator
	Liu, Daofu 4; Chen, Tianshi 4; Liu, Shaoli 4; Zhou, Jinhong 3; Zhou, Shengyuan 4; Teman, Olivier 1; Feng, Xiaobing 4; Zhou, Xuehai 3; Chen, Yunji 2
刊名	ACM SIGPLAN NOTICES
	2015-04-01
卷号	50 期号:4 页码:369-381
ISSN号	0362-1340
DOI	10.1145/2694344.2694358
英文摘要	Machine Learning (ML) techniques are pervasive tools in various emerging commercial applications, but have to be accommodated by powerful computer systems to process very large data. Although general-purpose CPUs and GPUs have provided straightforward solutions, their energy-efficiencies are limited due to their excessive supports for flexibility. Hardware accelerators may achieve better energy-efficiencies, but each accelerator often accommodates only a single ML technique (family). According to the famous No-Free-Lunch theorem in the ML domain, however, an ML technique performs well on a dataset may perform poorly on another dataset, which implies that such accelerator may sometimes lead to poor learning accuracy. Even if regardless of the learning accuracy, such accelerator can still become inapplicable simply because the concrete ML task is altered, or the user chooses another ML technique. In this study, we present an ML accelerator called PuDianNao, which accommodates seven representative ML techniques, including k-means, k-nearest neighbors, naive bayes, support vector machine, linear regression, classification tree, and deep neural network. Benefited from our thorough analysis on computational primitives and locality properties of different ML techniques, PuDianNao can perform up to 1 0 5 6 GOP/s (e.g., additions and multiplications) in an area of 3 : 5 1 mm 2, and consumes 596 mW only. Compared with the NVIDIA K20M GPU (28nm process), PuDianNao (65nm process) is 1.20x faster, and can reduce the energy by 128.41x.
资助项目	NSF of China[61100163] ; NSF of China[61133004] ; NSF of China[61222204] ; NSF of China[61221062] ; NSF of China[61303158] ; NSF of China[61473275] ; NSF of China[61432016] ; NSF of China[61472396] ; 973 Program of China[2015CB358800] ; 973 Program of China[2011CB302500] ; Strategic Priority Research Program of the CAS[XDA06010403] ; International Collaboration Key Program of the CAS[171111KYSB20130002] ; Google Faculty Research Award ; Intel Collaborative Research Institute for Computational Intelligence (ICRI-CI) ; 10,000 talent program ; 1,000 talent program
WOS研究方向	Computer Science
语种	英语
出版者	ASSOC COMPUTING MACHINERY
WOS记录号	WOS:000370874900026
内容类型	期刊论文
源URL	[http://119.78.100.204/handle/2XEOYT63/8829]
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Chen, Tianshi
作者单位	1.Inria, Villers, France 2.ICT, SKLCA, CAS Ctr Excellence Brain Sci, Beijing, Peoples R China 3.USTC, Hefei, Peoples R China 4.ICT, SKLCA, Beijing, Peoples R China
推荐引用方式 GB/T 7714	Liu, Daofu,Chen, Tianshi,Liu, Shaoli,et al. PuDianNao: A Polyvalent Machine Learning Accelerator[J]. ACM SIGPLAN NOTICES,2015,50(4):369-381.
APA	Liu, Daofu.,Chen, Tianshi.,Liu, Shaoli.,Zhou, Jinhong.,Zhou, Shengyuan.,...&Chen, Yunji.(2015).PuDianNao: A Polyvalent Machine Learning Accelerator.ACM SIGPLAN NOTICES,50(4),369-381.
MLA	Liu, Daofu,et al."PuDianNao: A Polyvalent Machine Learning Accelerator".ACM SIGPLAN NOTICES 50.4(2015):369-381.