Dresden 2017 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
DS: Fachverband Dünne Schichten
DS 36: Postersession I
DS 36.7: Poster
Mittwoch, 22. März 2017, 17:00–19:00, P2-EG
BFO based memristor as artificial synapse in machine learning circuits — •Mahdi Kiani1, Nan Nu1, Christian Mayr2, Danio Bürger1, Iluna Skrupa1,3, Stefan Shultz4, Oliver Schmidt1,5, and Heidemarie Schmidt1,4 — 1Materials systems for nanoelectronics, chemnitz university of technology, chemnitz, germany — 2Highly-parallel vlsi systems and neuro-microelectronics, technische universität dresden, dresden, germany — 3Helmholtz-zentrum dresden-rossendorf, institute of ion beam Physics and materials research,dresden, germany — 4Fraunhofer-Institut für Elektronische Nanosysteme, Abteilung Back-End of Line, Technologie, Chemnitz, Germany — 5Institute for Integrative Nanosciences, IFW Dresden, Dresden, Germany
Neuromorphic engineering takes advantage of artificial neurons and artificial synapses to mimic the most complicated human attribute, learning. BiFeO3 memristor as artificial synapse and newly designed electronic circuit as artificial neuron are used to implement associative, supervised, unsupervised, and deep learning. Spike-timing dependent plasticity (STDP) with a single pairing of one presynaptic voltage spike and one postsynaptic voltage spike with different time delays for Long-term potentiation (LTP) which determine learning status and Long-term depression (LTD) where forgetting occurs and also number cycle dependent plasticity (NCDP) for both student and teacher artificial synapses are demonstrated