中国科学院机构知识库网格
Chinese Academy of Sciences Institutional Repositories Grid
首页 机构 成果 学者
  
  1. CAS IR Grid
  2. 计算技术研究所
  3. 中国科学院计算技术研究所
  4. 中国科学院计算技术研究所期刊论文
  5. 英文
Breaking the Interaction Wall: A DLPU-Centric Deep Learning Computing System

文献类型：期刊论文

作者Du, Zidong4,5; Guo, Qi5; Zhao, Yongwei4,5; Zeng, Xi3,4,5; Li, Ling2; Cheng, Limin2; Xu, Zhiwei3,5; Sun, Ninghui3,5; Chen, Yunji1,3,5
刊名IEEE TRANSACTIONS ON COMPUTERS
出版日期2022
卷号71期号:1页码:209-222
关键词Deep learning Central Processing Unit Process control Task analysis Computational modeling Pipelines Runtime Neural net accelerators system architectures interaction wall
ISSN号0018-9340
DOI10.1109/TC.2020.3044245
英文摘要Due to the broad successes of deep learning, many CPU-centric artificial intelligent computing systems employ specialized devices such as GPUs, FPGAs, and ASICs, which can be named as Deep Learning Processing Units (DLPUs), for processing computation-intensive deep learning tasks. The separation between the scalar control operations mapped on CPUs and the vector computation operations mapped on DLPUs causes the frequent and costly interactions between CPUs and DLPUs, leading to the Interaction Wall. Moreover, the increasing algorithm complexity and DLPU computation speed would further aggravate the interaction wall substantially. To break the interaction wall, we propose a novel DLPU-centric deep learning computing system consisting of an exception-oriented programming (EOP) model and the architectural support of CPULESS DLPU. The EOP model processes scalar control operations of a deep learning task as exception handlers to maximally avoid stalling the crucial and dominated vector computation operations. Together with the CPULESS DLPU which integrates a scalar processing unit (SPU) for scalar control operations and the parallel processing unit (PPU) for vector computation operations into a fused pipeline, the proposed DLPU-centric system can cost-effectively leverage the EOP model to execute the two kinds of operations simultaneously without disturbing each other. Compared with a state-of-the-art commodity CPU-centric system with discrete V100 GPU via PCIe bus, experimental results show that our DLPU-centric system achieves 10.30x better performance and 92.99 percent energy savings, respectively. Moreover, compared with a CPU-centric version of DLPU system where the SPU serves as the host with integrated PPU, the proposed DLPU-centric system still achieves 15.60 percent better performance from avoided interactions.
资助项目National Key Research and Development Program of China[2018AAA0103300] ; National Key Research and Development Program of China[2017YFA0700900] ; National Key Research and Development Program of China[2017YFA0700902] ; National Key Research and Development Program of China[2017YFA0700901] ; National Key Research and Development Program of China[2019AAA0103802] ; National Key Research and Development Program of China[2020AAA0103802] ; NSF of China[61532016] ; NSF of China[61732007] ; NSF of China[61672491] ; NSF of China[61732002] ; NSF of China[61925208] ; NSF of China[61702478] ; NSF of China[61732020] ; NSF of China[61906179] ; NSF of China[U19B2019] ; NSF of China[U20A20227] ; Beijing Natural Science Foundation[JQ18013] ; Key Research Projects in Frontier Science of Chinese Academy of Sciences[QYZDB-SSW-JSC001] ; Strategic Priority Research Program of Chinese Academy of Science[XDB32050200] ; Strategic Priority Research Program of Chinese Academy of Science[XDC01020000] ; Strategic Priority Research Program of Chinese Academy of Science[XDC05010300] ; Strategic Priority Research Program of Chinese Academy of Science[XDC08040102] ; Youth Innovation Promotion Association CAS ; Beijing Academy of Artificial Intelligence (BAAI) ; Beijing Nova Program of Science and Technology[Z191100001119093] ; Guangdong Science and Technology Program[2019B090909005]
WOS研究方向Computer Science ; Engineering
语种英语
WOS记录号WOS:000730414800017
出版者IEEE COMPUTER SOC
源URL[http://119.78.100.204/handle/2XEOYT63/17973]  
专题中国科学院计算技术研究所期刊论文_英文
通讯作者Zhao, Yongwei
作者单位1.Shanghai Res Ctr Brain Sci & Brain Inspired Intel, Inst Brain Intelligence Technol, Zhangjiang Lab BIT ZJLab, CAS Ctr Excellence Brain Sci & Intelligence Techn, Beijing 200031, Peoples R China
2.Chinese Acad Sci, Inst Software, Beijing 100190, Peoples R China
3.Univ Chinese Acad Sci, Beijing 100049, Peoples R China
4.Cambricon Technol, Beijing 100191, Peoples R China
5.Chinese Acad Sci, Inst Comp Technol, State Key Lab Comp Architecture, Beijing 100190, Peoples R China
推荐引用方式
GB/T 7714		 Du, Zidong,Guo, Qi,Zhao, Yongwei,et al. Breaking the Interaction Wall: A DLPU-Centric Deep Learning Computing System[J]. IEEE TRANSACTIONS ON COMPUTERS,2022,71(1):209-222.
APA Du, Zidong.,Guo, Qi.,Zhao, Yongwei.,Zeng, Xi.,Li, Ling.,...&Chen, Yunji.(2022).Breaking the Interaction Wall: A DLPU-Centric Deep Learning Computing System.IEEE TRANSACTIONS ON COMPUTERS,71(1),209-222.
MLA Du, Zidong,et al."Breaking the Interaction Wall: A DLPU-Centric Deep Learning Computing System".IEEE TRANSACTIONS ON COMPUTERS 71.1(2022):209-222.

入库方式： OAI收割

来源：计算技术研究所

浏览0
下载0
收藏0
其他版本

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。