Oxide-based analog synapse: Physical modeling, experimental characterization, and optimization

Analog switching in oxide synaptic device has been recently proposed as an important technology for realizing hardware neural network with online training ability. This paper develops a new physical model to quantify the analog weight modulation behaviors in the oxide-based analog synapse. The analog SET, RESET, and retention loss processes are simulated and verified by the experimental data measured from the fabricated HfOx based synapse. Based on the simulation results, key material parameters are captured, and optimization guidelines are provided.

[1]  Farnood Merrikh-Bayat,et al.  Training and operation of an integrated neuromorphic network based on metal-oxide memristors , 2014, Nature.

[2]  Peng Huang,et al.  Optimized learning scheme for grayscale image recognition in a RRAM based analog neuromorphic system , 2015, 2015 IEEE International Electron Devices Meeting (IEDM).

[3]  Shimeng Yu,et al.  On the stochastic nature of resistive switching in metal oxide RRAM: Physical modeling, monte carlo simulation, and experimental characterization , 2011, 2011 International Electron Devices Meeting.

[4]  P. Cochat,et al.  Et al , 2008, Archives de pediatrie : organe officiel de la Societe francaise de pediatrie.

[5]  D. Ielmini.  Filamentary-switching model in RRAM for time, energy and scaling projections , 2011, 2011 International Electron Devices Meeting.

[6]  B. DeSalvo,et al.  Microscopic understanding of the low resistance state retention in HfO2 and HfAlO based RRAM , 2014, 2014 IEEE International Electron Devices Meeting.

[7]  X. A. Tran,et al.  Oxide-based RRAM: Unified microscopic principle for both unipolar and bipolar switching , 2011, 2011 International Electron Devices Meeting.

[8]  Jinfeng Kang,et al.  A physical based analytic model of RRAM operation for circuit simulation , 2012, 2012 International Electron Devices Meeting.

[9]  Shimeng Yu,et al.  A neuromorphic visual system using RRAM synaptic devices with Sub-pJ energy and tolerance to variability: Experimental characterization and large-scale modeling , 2012, 2012 International Electron Devices Meeting.

[10]  Tuo-Hung Hou,et al.  3D synaptic architecture with ultralow sub-10 fJ energy per spike for neuromorphic computation , 2014, 2014 IEEE International Electron Devices Meeting.

[11]  Jongin Kim,et al.  Oxide based nanoscale analog synapse device for neural signal recognition system , 2015, 2015 IEEE International Electron Devices Meeting (IEDM).