Affordable Access

Oscillatory Neural Networks for Edge AI Computing

Authors
  • Delacour, Corentin
  • Carapezzi, Stefania
  • Abernot, Madeleine
  • Boschetto, Gabriele
  • Azemard, Nadine
  • Salles, Jérémie
  • Gil, Thierry
  • Todri-Sanial, Aida
Publication Date
Jul 07, 2021
Source
HAL-INRIA
Keywords
Language
English
License
Unknown
External links

Abstract

In this paper, we showcase the innovative concept of implementing Oscillatory Neural Networks (ONNs) for neuromorphic computing with beyond CMOS devices based on vanadium dioxide to mimic neurons and resistors to emulate synapses. We explore ONN technology potentials from device to analog circuit-level simulations. We report that ONN behaves like an associative memory and can implement energy-based models such as Hopfield Neural Networks on edge devices. Finally, as a proof of concept, a reconfigurable digital ONN is implemented on FPGA for pattern recognition tasks.

Report this publication

Statistics

Seen <100 times