Optical computing based on phase change materials
New and exciting approaches and architecture designs are transforming the world of computing. Photonic computing represents a solution with many advantages over traditional semiconductor systems, including higher transmission speeds, lower energy consumption and systems miniaturisation. Widespread implementation of photonic computing requires the development of efficient optical transistors and non-volatile optical memory.
Based on previously developed micro optical switches and all-photonic memory based on phase change materials (PCM), a collaboration of researchers at the University of Oxford has developed systems and techniques to utilise these components effectively in future optical processing and memory units. These developments help pave the way towards optical memcomputing devices (devices that carry out both data processing and storage) and creating “photonic synapses” and “optical neurons”.
Our information age depends on electronic devices, which exploit the flow of electrons to transmit information. However, increasing pressure to improve the speed and storage capacity of new devices has led to the quest for different methods. In this context, optical devices represent an attractive alternative for the management of information. It has been proposed that by using photons for computation, instead of electrons, higher bandwidth devices could be developed.
New optical solutions
Researchers from the University of Oxford have developed a novel array of new optical computing solutions based on previously developed optical memory and switching components using phase change materials.
These solutions include:
- Methods and devices for performing scalar, vector and matrix multiplications using optical computing components.
- A method for switching optical memories between volatile and non-volatile modes of use.
- A novel optical computing-equivalent of a digital-to-analogue converter.
- Novel techniques for reading, writing and resetting of optical memory elements.
Proof-of-concept demonstrators of the new optical computing solutions have been developed and initial testing work has been completed.
The developed solutions help to pave the way towards optical memcomputing devices (devices that carry out both data processing and storage) and creating “photonic synapses” and “optical neurons”.
These devices could form integral components in next generation artificial neural networks, general processing units and machine learning and artificial intelligence systems.
Several patent applications concerning the technology have been filed and Oxford University Innovation is actively seeking commercial partners to help develop the technology further and take an exclusive market position in relation to it.
about this technology