Regensburg 2016 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
HL: Fachverband Halbleiterphysik
HL 15: Quantum Dots and Wires: Fabrication and Devices
HL 15.8: Vortrag
Montag, 7. März 2016, 17:15–17:30, H16
An electro-photo-sensitive memristor for neuromorphic and arithmetic computing applications — •Patrick Maier1, Fabian Hartmann1, Monika Emmerling1, Christian Schneider1, Martin Kamp1, Sven Höfling1,2, and Lukas Worschech1 — 1Technische Physik and Wilhelm Conrad Röntgen Research Center for Complex Material Systems, Physikalisches Institut, Universität Würzburg, Am Hubland, D-97074 Würzburg, Germany — 2SUPA, School of Physics and Astronomy, University of St. Andrews, St. Andrews, KY16 9SS, UK
Memristors have state- and time-dependent resistances that allow the emulation of synaptic functionalities which are essential for learning and memory. We report the realization of a memristor on the mature III-V-semiconductor platform and present an electro-optical control of the conductance by tuning the amount of localized charge on a quantum dot floating gate. In analogy to synaptic strength modifications in neural networks, we show that the conductance can be increased (potentiation) or decreased (depression) by tuning the time difference between incoming voltage pulses. In addition, the localized charge on the quantum dot shifts the threshold voltages for the induction of potentiation and depression. The dependency of the threshold voltage for potentiation on the initial state in combination with the optical control of the conductance enables arithmetic computing applications of low power light pulses. Our findings may pave the way to the realization of electro-optical, memristor-based artificial neural networks with a memory-dependent ability of learning.