The World's First Ultra-Fast Photonic Information Processor

Photonic Computing
Photonic Computing

Researchers at the University of Oxford have discovered a method that uses the polarization of light to increase information storage density and computing performance using nanowires, according to a paper published June 15 in Science Advances.

Different wavelengths of light do not interact with each other. This is an exploitable feature used by fiber optics to transmit parallel data streams. The different polarizations of light do not interact with each other in the same way. Each polarization can act as its own information channel, allowing more data to be stored in multiple channels and a significant increase in information density.

June Sang Lee of the University of Oxford said: “The advantage of photonics over electronics is that light, as we all know, is faster and more useful at wide bandwidths. Ultimately, our goal was to take full advantage of photonics with tunable materials to enable faster and more intensive data processing.”

In collaboration with Professor C. David Wright of the University of Exeter, the research team developed a HAD (hybridized-active-dielectric) nanowire. The technique was to use a hybrid glassy material that exhibits switchable material properties upon illumination of optical pulses. Each nanowire showed selective responses to a particular polarization direction, so information can be processed simultaneously using multiple polarizations in different directions.

The World's First Ultra-Fast Photonic Information Processor
Hybrid nanowires that can selectively switch devices depending on polarization. Credits: June Sang Lee, Materials Department, University of Oxford

Using this concept, the researchers developed the first photonic computing processor to use polarizations of light.

Compared to typical electronic chips, photonic computing uses several polarization channels to increase the computational density by several orders of magnitude. Because these nanowires are replaced by nanosecond optical pulses, computation rates are faster.

The method called “Moore's Law” has been a method used to maximize computing density since the creation of the first integrated circuit in 1958. However, Artificial Intelligence and Machine Learning demand specialized hardware that is starting to push the boundaries of traditional systems. The most pressing issue in this field of computer engineering is “How can we pack more capability in a single transistor?” has been.

Professor Harish Bhaskaran's lab at Oxford University's Department of Materials has been exploring the use of light as a computing medium for more than a decade.

Project leader Professor Bhaskaran said: “This is just the beginning of what we want to see in the future.

Using all the degrees of freedom that light offers, including polarization, is our goal method to significantly parallelize information processing. It's certainly early stage research, but these are really fascinating ideas combining electronics, nonlinear materials and computation. There are so many exciting opportunities to work on it, which is always a good thing.”

source: phys.org

Similar Ads

Be the first to comment

your comment