skip to main content

DOE PAGESDOE PAGES

Title: Learning through ferroelectric domain dynamics in solid-state synapses

In the brain, learning is achieved through the ability of synapses to reconfigure the strength by which they connect neurons (synaptic plasticity). In promising solid-state synapses called memristors, conductance can be finely tuned by voltage pulses and set to evolve according to a biological learning rule called spike-timing-dependent plasticity (STDP). Future neuromorphic architectures will comprise billions of such nanosynapses, which require a clear understanding of the physical mechanisms responsible for plasticity. Here we report on synapses based on ferroelectric tunnel junctions and show that STDP can be harnessed from inhomogeneous polarization switching. Through combined scanning probe imaging, electrical transport and atomic-scale molecular dynamics, we demonstrate that conductance variations can be modelled by the nucleation-dominated reversal of domains. Finally, based on this physical model, our simulations show that arrays of ferroelectric nanosynapses can autonomously learn to recognize patterns in a predictable way, opening the path towards unsupervised learning in spiking neural networks.
Authors:
 [1] ;  [2] ;  [3] ; ORCiD logo [4] ; ORCiD logo [5] ;  [2] ;  [6] ;  [2] ;  [2] ;  [7] ;  [3] ;  [4] ;  [2] ;  [2] ;  [3] ;  [2]
  1. Univ. Paris-Saclay, Palaiseau (France); ETH Zurich, Zurich (Switzerland)
  2. Univ. Paris-Saclay, Palaiseau (France)
  3. Univ. of Bordeaux, Talence (France)
  4. Univ. of Arkansas, Fayetteville, AR (United States)
  5. Univ. Paris-Saclay, Orsay Cedex (France)
  6. Univ. Paris-Saclay, Palaiseau (France); Luxembourg Institute of Science and Technology (LIST), Belvaux (Luxembourg)
  7. Campus de l'Ecole Polytechnique, Palaiseau (France)
Publication Date:
Grant/Contract Number:
SC0002220; HR0011-15-2- 0038
Type:
Accepted Manuscript
Journal Name:
Nature Communications
Additional Journal Information:
Journal Volume: 8; Journal ID: ISSN 2041-1723
Publisher:
Nature Publishing Group
Research Org:
Univ. of Arkansas, Fayetteville, AR (United States)
Sponsoring Org:
USDOE Office of Science (SC), Basic Energy Sciences (BES) (SC-22); Defense Advanced Research Projects Agency (DARPA); European Union (EU); European Research Council (ERC)
Country of Publication:
United States
Language:
English
Subject:
60 APPLIED LIFE SCIENCES; electronic devices; ferroelectrics and multiferroics
OSTI Identifier:
1367544

Boyn, Soren, Grollier, Julie, Lecerf, Gwendal, Xu, Bin, Locatelli, Nicolas, Fusil, Stephane, Girod, Stephanie, Carretero, Cecile, Garcia, Karin, Xavier, Stephane, Tomas, Jean, Bellaiche, Laurent, Bibes, Manuel, Barthelemy, Agnes, Saighi, Sylvain, and Garcia, Vincent. Learning through ferroelectric domain dynamics in solid-state synapses. United States: N. p., Web. doi:10.1038/ncomms14736.
Boyn, Soren, Grollier, Julie, Lecerf, Gwendal, Xu, Bin, Locatelli, Nicolas, Fusil, Stephane, Girod, Stephanie, Carretero, Cecile, Garcia, Karin, Xavier, Stephane, Tomas, Jean, Bellaiche, Laurent, Bibes, Manuel, Barthelemy, Agnes, Saighi, Sylvain, & Garcia, Vincent. Learning through ferroelectric domain dynamics in solid-state synapses. United States. doi:10.1038/ncomms14736.
Boyn, Soren, Grollier, Julie, Lecerf, Gwendal, Xu, Bin, Locatelli, Nicolas, Fusil, Stephane, Girod, Stephanie, Carretero, Cecile, Garcia, Karin, Xavier, Stephane, Tomas, Jean, Bellaiche, Laurent, Bibes, Manuel, Barthelemy, Agnes, Saighi, Sylvain, and Garcia, Vincent. 2017. "Learning through ferroelectric domain dynamics in solid-state synapses". United States. doi:10.1038/ncomms14736. https://www.osti.gov/servlets/purl/1367544.
@article{osti_1367544,
title = {Learning through ferroelectric domain dynamics in solid-state synapses},
author = {Boyn, Soren and Grollier, Julie and Lecerf, Gwendal and Xu, Bin and Locatelli, Nicolas and Fusil, Stephane and Girod, Stephanie and Carretero, Cecile and Garcia, Karin and Xavier, Stephane and Tomas, Jean and Bellaiche, Laurent and Bibes, Manuel and Barthelemy, Agnes and Saighi, Sylvain and Garcia, Vincent},
abstractNote = {In the brain, learning is achieved through the ability of synapses to reconfigure the strength by which they connect neurons (synaptic plasticity). In promising solid-state synapses called memristors, conductance can be finely tuned by voltage pulses and set to evolve according to a biological learning rule called spike-timing-dependent plasticity (STDP). Future neuromorphic architectures will comprise billions of such nanosynapses, which require a clear understanding of the physical mechanisms responsible for plasticity. Here we report on synapses based on ferroelectric tunnel junctions and show that STDP can be harnessed from inhomogeneous polarization switching. Through combined scanning probe imaging, electrical transport and atomic-scale molecular dynamics, we demonstrate that conductance variations can be modelled by the nucleation-dominated reversal of domains. Finally, based on this physical model, our simulations show that arrays of ferroelectric nanosynapses can autonomously learn to recognize patterns in a predictable way, opening the path towards unsupervised learning in spiking neural networks.},
doi = {10.1038/ncomms14736},
journal = {Nature Communications},
number = ,
volume = 8,
place = {United States},
year = {2017},
month = {4}
}

Works referenced in this record:

A fast, high-endurance and scalable non-volatile memory device made from asymmetric Ta2O5−x/TaO2−x bilayer structures
journal, July 2011
  • Lee, Myoung-Jae; Lee, Chang Bum; Lee, Dongsoo
  • Nature Materials, Vol. 10, Issue 8, p. 625-630
  • DOI: 10.1038/nmat3070

Memristive devices for computing
journal, January 2013
  • Yang, J. Joshua; Strukov, Dmitri B.; Stewart, Duncan R.
  • Nature Nanotechnology, Vol. 8, Issue 1, p. 13-24
  • DOI: 10.1038/nnano.2012.240