RRAM based Synaptic Devices for Neuromorphic Visual Systems

Kang, J. F.1; Gao, B.1; Huang, P.1; Liu, L. F.1; Liu, X. Y.1; Yu, H. Y.2; Yu, S.3; Wong, H. -S. Philip4

10.1109/ICDSP.2015.7252074

2015

2015 IEEE INTERNATIONAL CONFERENCE ON DIGITAL SIGNAL PROCESSING (DSP)

2015-September

1219-1222

Singapore, Singapore

345 E 47TH ST, NEW YORK, NY 10017 USA

IEEE

Neuromorphic computing is an attractive computation paradigm with the features of massive parallelism, adaptivity to the complex input information, and tolerance to errors. As one of the most crucial components in a neuromorphic system, the electronic synapse requires high device integration density and low-energy consumption. Oxide-based resistive switching devices (RRAM) have emerged as the leading candidate to realize the synapse functions due to the extra-low energy loss per spike. This work will address the design and optimization of oxide-based RRAM synaptic devices and the impacts of the synaptic devices parameters on the performance of neuromorphic visual system. Possible solutions are also provided to suppress the intrinsic variation of the oxide-RRAM based synaptic devices to achieve high recognition accuracy and efficiency of neuromorphic visual systems

RRAM Neuromorphic Computing Synapse Neural Cell

其他

英语

CPCI-S ; EI

Engineering

Engineering, Electrical & Electronic

WOS:000380506600256

20160501876956

Energy dissipation ; Energy utilization ; RRAM

Energy Utilization:525.3 ; Energy Losses (industrial and residential):525.4

Web of Science

1.Peking Univ, Inst Microelect, Beijing, Peoples R China
2.South Univ Sci & Technol China, Dept EEE, Shenzhen 518055, Peoples R China
3.Arizona State Univ, Sch Elect Comp & Energy Engn, Tempe, AZ USA
4.Stanford Univ, Dept Elect Engn, Stanford, CA 94305 USA

Kang, J. F.,Gao, B.,Huang, P.,et al. RRAM based Synaptic Devices for Neuromorphic Visual Systems[C]. 345 E 47TH ST, NEW YORK, NY 10017 USA:IEEE,2015:1219-1222.