Scientists unveil fastest optical neuromorphic processor

An international team of researchers led by Swinburne University of Technology have demonstrated a fast and powerful optical neuromorphic processor for artificial intelligence, which operates faster than 10 trillion operations per second (TeraOPs/s) and is capable of processing ultralarge-scale data.

Published in the journal Nature, this breakthrough represents a leap forward for neural networks and neuromorphic processing.

Artificial neural networks, a key form of AI, can ‘learn’ and perform complex operations with applications to computer vision, natural language processing, facial recognition, speech translation, playing strategy games, medical diagnosis and other areas. Artificial neural networks extract key features of raw data to predict properties and behaviour with accuracy and simplicity.

Led by Swinburne’s Professor David Moss, Dr Xingyuan (Mike) Xu from Monash University and Distinguished Professor Arnan Mitchell from RMIT University, the international team achieved a feat in optical neural networks: dramatically accelerating computer speed and processing power.

The researchers demonstrated an optical neuromorphic processor operating more than 1000 times faster than any previous processor, with the system also processing ultralarge-scale images — enough to achieve full facial recognition, which other processors have been unable to accomplish.

“This breakthrough was achieved with ‘optical micro-combs’, as was our world-record internet data speed reported in May 2020,” said Professor Moss, Director of Swinburne’s Optical Sciences Centre.

While the Google TPU electronic processor can operate beyond 100 TeraOPs/s, this is achieved with tens of thousands of parallel processors. In contrast, the team’s optical system uses a single processor and was achieved using a new technique of simultaneously interleaving the data in time, wavelength and spatial dimensions through an integrated micro-comb source.

Microcombs are devices that act like a rainbow made up of hundreds of high-quality infrared lasers on a single chip. They are faster, smaller, lighter and cheaper than other optical sources.

“In the 10 years since I co-invented them, integrated micro-comb chips have become enormously important and it is truly exciting to see them enabling these huge advances in information communication and processing. Micro-combs offer enormous promise for us to meet the world’s insatiable need for information,” said Professor Moss.

Dr Xu, who served as co-lead author of the study, said the processor can serve as a universal ultrahigh-bandwidth front end for any neuromorphic hardware — optical or electronic based — bringing massive-data machine learning for real-time ultrahigh-bandwidth data within reach.

“We’re currently getting a sneak peek of how the processors of the future will look. It’s really showing us how dramatically we can scale the power of our processors through the innovative use of microcombs,” said Dr Xu.

Professor Mitchell said the technology is applicable to all forms of processing and communications, with the research team hoping to realise fully integrated systems on a chip, thereby reducing cost and energy consumption.

Professor Damien Hicks, from Swinburne University and the Walter and Eliza Hall Institute, said this breakthrough shows how a new optical technology makes convolutional neural networks faster and more efficient, and is a demonstration of the benefits of cross-disciplinary thinking.

“Convolutional neural networks have been central to the artificial intelligence revolution, but existing silicon technology increasingly presents a bottleneck in processing speed and energy efficiency,” said Professor Hicks.

Image credit: Swinburne University of Technology.