New electro-optical device may herald faster processors

Researchers have created a novel device that bridges the fields of optical and electronic computing, an advance that may lead to faster and more energy-efficient memories and processors.

The researchers, including those from the University of Oxford in the UK, said until now a compact device using light to transfer as well as encode information could not be easily interfaced with the architecture of traditional electronic computers.

This was because electrical chips, in which the subatomic particles electrons flow through, needed to be small to operate efficiently, whereas optical chips -- with the light particles photons flowing through them -- are required to be large.

They explained this was because the wavelength of light is larger than that of electrons.

As part of the study, published in the journal Science Advances, the scientists came up with a solution to confine light into nanoscopic dimensions, and have developed devices with dual electrical-optical functionality.

The size reduction and a significantly increased energy density allowed them to bridge the incompatibility of photons and electrons for data storage and computation, the researchers said.

They showed that by sending either electrical or optical signals, the state of a photo- and electro-sensitive material is transformed between two different states.

The state of this phase-transforming material was read out by either light or electronics, meaning memory could be encoded and retrieved in the form of these states.

This makes the device the first electro-optical nanoscale memory cell without any volatile characteristics, the researchers said.

"This is a very promising path forward in computation and especially in fields where high processing efficiency is needed," said study co-author Nikolaos Farmakidis from the University of Oxford.

"This naturally includes artificial intelligence applications wherein many occasions the needs for high-performance, low-power computing far exceeds our current capabilities," said Nathan Youngblood, co-author of the study, also from the University of Oxford.

The researchers said interfacing light-based photonic computing with its electrical counterpart may be the key to the next chapter in integrated circuits and other semiconductor applications.