Project

High-speed low-power neuromorphic photonic information processing with chaotic cavities

Code
3G022520
Duration
01 January 2020 → 31 December 2023
Funding
Research Foundation - Flanders (FWO)
Research disciplines
  • Engineering and technology
    • Pattern recognition and neural networks
Keywords
photonic neuromorphic computing
 
Project description

The human brain still outperforms digital computers in its flexibility and in its performance on pattern recognition tasks like image and speech recognition Moreover, it is able to achieve this at very modest power consumption levels, typically the power equivalent to a light bulb, whereas (super)computers require orders of magnitude more power In this project, we want to take inspiration from the brain to build a novel class of computing system that is not only bio-inspired, but also uses light as the fundamental carrier of information, as opposed to electricity in current computers Compared to electronics, photonics has the advantage that many physical effects can be extremely fast Moreover, we will build our systems using the same mature mass-fabrication silicon technology that is used for computer chips, but now building chips that operate on light All of this opens up the potential to solve certain classes of problems (routing of internet signals, processing of medical images, ) at very high speed, low power consumption and low cost