EnGN: A High-Throughput and Energy-Efficient Accelerator for Large Graph Neural Networks

doi:10.1109/TC.2020.3014632

	EnGN: A High-Throughput and Energy-Efficient Accelerator for Large Graph Neural Networks
	Liang, Shengwen 1,2; Wang, Ying 1,2; Liu, Cheng 1,2; He, Lei 1,2; Li, Huawei 1,2; Xu, Dawen 1,2; Li, Xiaowei 1,2
	2021-09-01
发表期刊	IEEE TRANSACTIONS ON COMPUTERS
ISSN	0018-9340
卷号	70 期号:9 页码:1511-1525
摘要	Graph neural networks (GNNs) emerge as a powerful approach to process non-euclidean data structures and have been proved powerful in various application domains such as social networks and e-commerce. While such graph data maintained in real-world systems can be extremely large and sparse, thus employing GNNs to deal with them requires substantial computational and memory overhead, which induces considerable energy and resource cost on CPUs and GPUs. In this article, we present a specialized accelerator architecture, EnGN, to enable high-throughput and energy-efficient processing of large-scale GNNs. The proposed EnGN is designed to accelerate the three key stages of GNN propagation, which is abstracted as common computing patterns shared by typical GNNs. To support the key stages simultaneously, we propose the ring-edge-reduce(RER) dataflow that tames the poor locality of sparsely-and-randomly connected vertices, and the RER PE-array to practice RER dataflow. In addition, we utilize a graph tiling strategy to fit large graphs into EnGN and make good use of the hierarchical on-chip buffers through adaptive computation reordering and tile scheduling. Overall, EnGN achieves performance speedup by 1802.9X, 19.75X, and 2.97X and energy efficiency by 1326.35X, 304.43X, and 6.2X on average compared to CPU, GPU, and a state-of-the-art GCN accelerator HyGCN, respectively.
关键词	Neural networks Hardware System-on-chip Task analysis Feature extraction Memory management Graph neural network accelerator architecture hardware acceleration
DOI	10.1109/TC.2020.3014632
收录类别	SCI
语种	英语
资助项目	National Key Research and Development Program of China[2018AAA0102700] ; National Natural Science Foundation of China[61874124] ; National Natural Science Foundation of China[61876173] ; National Natural Science Foundation of China[61532017] ; National Natural Science Foundation of China[61772300] ; National Natural Science Foundation of China[61902375] ; Strategic Priority Research Program of Chinese Academy of Sciences[XDC05030201] ; Strategic Priority Research Program of Chinese Academy of Sciences[YESS2016qnrc001]
WOS研究方向	Computer Science ; Engineering
WOS类目	Computer Science, Hardware & Architecture ; Engineering, Electrical & Electronic
WOS记录号	WOS:000682123200015
出版者	IEEE COMPUTER SOC
引用统计	被引频次：89[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/17388
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Wang, Ying; Li, Huawei
作者单位	1.Chinese Acad Sci, State Key Lab Comp Architecture, Inst Comp Technol, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, Beijing 100190, Peoples R China
推荐引用方式 GB/T 7714	Liang, Shengwen,Wang, Ying,Liu, Cheng,et al. EnGN: A High-Throughput and Energy-Efficient Accelerator for Large Graph Neural Networks[J]. IEEE TRANSACTIONS ON COMPUTERS,2021,70(9):1511-1525.
APA	Liang, Shengwen.,Wang, Ying.,Liu, Cheng.,He, Lei.,Li, Huawei.,...&Li, Xiaowei.(2021).EnGN: A High-Throughput and Energy-Efficient Accelerator for Large Graph Neural Networks.IEEE TRANSACTIONS ON COMPUTERS,70(9),1511-1525.
MLA	Liang, Shengwen,et al."EnGN: A High-Throughput and Energy-Efficient Accelerator for Large Graph Neural Networks".IEEE TRANSACTIONS ON COMPUTERS 70.9(2021):1511-1525.