Replacing lasers with less complex light sources can improve performance in some optical applications, researchers from the University of Oxford have found.

The study, which involved collaboration with the Universities of Muenster, Heidelberg and Ghent, indicated applications which generally require high-spec lasers could instead use cheaper, less energy-consuming alternatives.

A light source's properties are often characterised using a measure known as 'coherence', which refers to how consistent the light waves are with each other over time and space.

Low coherence light sources, such as the sun and light bulbs, emit light in a wide range of colour, or wavelengths.

On the other hand, high quality engineered light sources, such as lasers, have a very narrow wavelength range and typically appear as a single colour.

It has been assumed that using more coherent light sources enhances system performance and device functionalities, for instance, by enabling higher resolution and more precise measurements.

This assumption served as the foundation for numerous modern applications including optical communications and medical imaging technologies.

However, the researchers have discovered low coherence light sources can surprisingly outperform lasers in certain scenarios, challenging this previous assumption.

This proved to be the case in photonic AI accelerators, an emerging technology which uses photons instead of electrons to conduct AI computations.

The team used a partially coherent light source and distributed it evenly into various input channels for a parallel AI computational array.

A test carried out using this system identified Parkinson’s disease patients by analysing their walk, with over 92 per cent classification accuracy.

A system using only one partially coherent light source with nine input channels could perform high-speed AI tasks at a speed of 100 billion operations per second.

This would normally require the use of multiple separate coherent lasers in a coherent photonic AI accelerator.

Dr Bowei Dong from the Department of Materials, University of Oxford, is first author of the study.

He said: "The benefit of using 'poorer' light sources has a scaling effect.

"You can run your AI models 100 times faster compared to a laser system, if the photonic accelerator scales to 100 input channels."

Professor Harish Bhaskaran from the Department of Materials, University of Oxford, is co-founder of Salience Labs, who led the work.

He added: "While this work showcases the use of such partially coherent light in some emerging areas of photonic computing, we will in future also investigate whether this insight might apply to optical communications, particularly in the emerging optical interconnect technology space.

"This is an area of research that is moving rapidly with a lot of interesting science and engineering to explore."