Dresden 2017 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
TT: Fachverband Tiefe Temperaturen
TT 11: SYCE: Novel Functionality and Topology-Driven Phenomena in Ferroics and Correlated Electron Systems (joint symposium DF, DS, KR, MA, MI, TT, organized by DS)
TT 11.5: Invited Talk
Monday, March 20, 2017, 17:30–18:00, HSZ 02
Learning through ferroelectric domain dynamics in solidstate synapses — Sören Boyn1, Gwendal Lecerf2, Stéphane Fusil1, Sylvain Saïghi2, Agnès Barthélémy1, Julie Grollier1, Vincent Garcia1, and •Manuel Bibes1 — 1Unité Mixte de Physique CNRS/Thales, Palaiseau FRANCE — 2IMS Laboratory, U. Bordeaux FRANCE
In the brain, learning is achieved through the ability of synapses to reconfigure the strength by which they connect two neurons. Artificial hardware with performances emulating those of biological systems require electronic nanosynapses endowed with such plasticity. Promising solid-state synapses are memristors, simple two-terminal nanodevices that can be finely tuned by voltage pulses. Their conductance evolves according to a learning rule called spike-timing-dependent plasticity, conjectured to underlie unsupervised learning in our brains. We will report on purely electronic ferroelectric synapses and show that spike timing-dependent plasticity can be harnessed and tuned from intrinsically inhomogeneous ferroelectric polarisation switching. Through combined scanning probe imaging and electrical transport experiments, we demonstrate that conductance variations in such BiFeO3-based ferroelectric memristors can be accurately controlled and modelled by the nucleation-dominated electric-feld switching of domains with different polarisations. Our results show that ferroelectric nanosynapses are able to learn in a reliable and predictable way, opening the way towards unsupervised learning in spiking neural networks.