Application of Deep Compression Technique in Spiking Neural Network Chip

被引:10
|
作者
Liu, Yanchen [1 ]
Qian, Kun [1 ]
Hu, Shaogang [1 ]
An, Kun [2 ]
Xu, Sheng [3 ]
Zhan, Xitong [1 ]
Wang, J. J. [1 ]
Guo, Rui [1 ]
Wu, Yuancong [1 ]
Chen, Tu-Pei [4 ]
Yu, Qi [5 ]
Liu, Yang [5 ]
机构
[1] Univ Elect Sci & Technol China, Chengdu 610054, Peoples R China
[2] Omnivis Thchnol Inc, Shanghai 201210, Peoples R China
[3] Cambricon Informat Technol Co, Shanghai, Peoples R China
[4] Nanyang Technol Univ, Sch Elect & Elect Engn, Singapore 639798, Singapore
[5] Univ Elect Sci & Technol China, State Key Lab Elect Thin Films & Intergrated, Chengdu 610054, Peoples R China
基金
中国国家自然科学基金;
关键词
Deep compression; network-on-chip; neuron; spiking neural network; synapse; ON-CHIP; DESIGN; SYSTEM;
D O I
10.1109/TBCAS.2019.2952714
中图分类号
R318 [生物医学工程];
学科分类号
0831 ;
摘要
In this paper, a reconfigurable and scalable spiking neural network processor, containing 192 neurons and 6144 synapses, is developed. By using deep compression technique in spiking neural network chip, the amount of physical synapses can be reduced to 1/16 of that needed in the original network, while the accuracy is maintained. This compression technique can greatly reduce the number of SRAMs inside the chip as well as the power consumption of the chip. This design achieves throughput per unit area of 1.1 GSOP/(s. mm(2)) at 1.2 V, and energy consumed per SOP of 35 pJ. A 2-layer fully-connected spiking neural network is mapped to the chip, and thus the chip is able to realize handwritten digit recognition on MNIST with an accuracy of 91.2%.
引用
收藏
页码:274 / 282
页数:9
相关论文
共 50 条
  • [1] A Spiking Neural Network Chip for Odor Data Classification
    Hsieh, Hung-Yi
    Tang, Kea-Tiong
    [J]. 2012 IEEE ASIA PACIFIC CONFERENCE ON CIRCUITS AND SYSTEMS (APCCAS), 2012, : 88 - 91
  • [2] DNNZip: Selective Layers Compression Technique in Deep Neural Network Accelerators
    Landhiri, Habiba
    Palesi, Maurizio
    Monteleone, Salvatore
    Patti, Davide
    Ascia, Giuseppe
    Lorandel, Jordane
    Bourdel, Emmanuelle
    Catania, Vincenzo
    [J]. 2020 23RD EUROMICRO CONFERENCE ON DIGITAL SYSTEM DESIGN (DSD 2020), 2020, : 526 - 533
  • [3] Spiking SiamFC++: deep spiking neural network for object tracking
    Shuiying Xiang
    Tao Zhang
    Shuqing Jiang
    Yanan Han
    Yahui Zhang
    Xingxing Guo
    Licun Yu
    Yuechun Shi
    Yue Hao
    [J]. Nonlinear Dynamics, 2024, 112 : 8417 - 8429
  • [4] A High Learning Capability Probabilistic Spiking Neural Network Chip
    Hsieh, Hung-Yi
    Li, Pin-Yi
    Yang, Cheng-Han
    Tang, Kea-Tiong
    [J]. 2018 INTERNATIONAL SYMPOSIUM ON VLSI DESIGN, AUTOMATION AND TEST (VLSI-DAT), 2018,
  • [5] An Analog Probabilistic Spiking Neural Network with On-Chip Learning
    Hsieh, Hung-Yi
    Li, Pin-Yi
    Tang, Kea-Tiong
    [J]. NEURAL INFORMATION PROCESSING (ICONIP 2017), PT VI, 2017, 10639 : 777 - 785
  • [6] Spiking Neural Network E-NOSE Classifier Chip
    Abdel-Aty-Zohdy, Hoda S.
    Allen, Jacob N.
    Ewing, Robert L.
    [J]. PROCEEDINGS OF THE IEEE 2010 NATIONAL AEROSPACE AND ELECTRONICS CONFERENCE (NAECON), 2010, : 374 - 378
  • [7] Hardware spiking neural network prototyping and application
    Cawley, Seamus
    Morgan, Fearghal
    McGinley, Brian
    Pande, Sandeep
    McDaid, Liam
    Carrillo, Snaider
    Harkin, Jim
    [J]. GENETIC PROGRAMMING AND EVOLVABLE MACHINES, 2011, 12 (03) : 257 - 280
  • [8] Hardware spiking neural network prototyping and application
    Seamus Cawley
    Fearghal Morgan
    Brian McGinley
    Sandeep Pande
    Liam McDaid
    Snaider Carrillo
    Jim Harkin
    [J]. Genetic Programming and Evolvable Machines, 2011, 12 : 257 - 280
  • [9] SpikeMS: Deep Spiking Neural Network for Motion Segmentation
    Parameshwara, Chethan M.
    Li, Simin
    Fermuller, Cornelia
    Sanket, Nitin J.
    Evanusa, Matthew S.
    Aloimonos, Yiannis
    [J]. 2021 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS), 2021, : 3414 - 3420
  • [10] An Unsupervised Spiking Deep Neural Network for Object Recognition
    Song, Zeyang
    Wu, Xi
    Yuan, Mengwen
    Tang, Huajin
    [J]. ADVANCES IN NEURAL NETWORKS - ISNN 2019, PT II, 2019, 11555 : 361 - 370