DPG Phi
Verhandlungen
Verhandlungen
DPG

Regensburg 2019 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

HL: Fachverband Halbleiterphysik

HL 12: HL Poster I

HL 12.61: Poster

Montag, 1. April 2019, 17:30–20:00, Poster E

From Hybrid Si Nanowire Transistors to Artificial Neurons — •Khrystyna Nych1, Eunhye Baek1,2, Taiuk Rim3, Larysa Baraban1,2, and Gianaurelio Cuniberti1,21Institute for Materials Science and Max Bergmann Center of Biomaterials, TU Dresden — 2Center for Advancing Electronics Dresden — 3Department of Electrical Engineering, Pohang University of Science and Technology

The rampant development of neuromorphic computing technologies inspired multiple alternatives to be presented - from complex CMOS circuits to novel circuit elements with external post-processing. Presently, both extensive training iterations and a large number of devices is necessary to enable error-free low pixel quantity image recognition tasks. We are striving to combine both strategies by proposing an electrically-controlled n-doped Si nanowire transistor on a silicon on insulator wafer as the computational means combined with an ionic silicate film spin-coated on top to emulate the biological function of the cell body. This cell body, also known as the soma, is responsible for processing multiple dendrite inputs and generating an electric signal, the action potential, which is sent through the axon to the next neuron. An electrical bias at the gate stimulates a polarization in the silicate film which modulates the nanowire current, depending on the voltage polarity. The time constants for charging and discharging the film depend on the concentration of ions in the sol-gel matrix, the intensity of the voltage applied, the frequency of the input pulses and the integration time.

100% | Mobil-Ansicht | English Version | Kontakt/Impressum/Datenschutz
DPG-Physik > DPG-Verhandlungen > 2019 > Regensburg