Parallel in-memory photonic computing using continuous-time data representation

Image from Licence Details: Parallel in-memory photonic computing using continuous-time data representation

Applications: Data communication, photonic computing, optical neural networks, machine learning, deep learning, edge computing

The parallelism enhancement from each of ‘continuous-time data representation’ (N) and ‘in-memory photonic computing’ (M/p) can be synergized to reach a higher MVM computation parallelism (N×M/p)

Features Benefits
  • Light is the data carrier rather than electric signals
  • Far more information can be transmitted in a given time frame – a THz data rate.
  • Continuous-time data representation
  • Using this instead if binary data representation allows for full benefits of the optical bandwidth to enhance data communication parallelism. It has been shown to embed 50 values in one time cycle.
  • Both continuous-time data representation methods from electronics and the photonic wavelength-division multiplexing are combined
  • This achieves a significantly higher number of simultaneous data streams
  • Large frequency difference between discrete optical frequencies
  • So each optical frequency can act as a carrier frequency without interference.


Request more information
about this technology

Back to Technologies Available

Sparks Background Image

Ready to get in touch?

Contact Us
Sparks Background Image
© Oxford University Innovation