An international team led by researchers at Swinburne University of Technology in Australia have developed a universal photonic optical vector convolutional accelerator operating at more than ten trillion operations per second (TOPS). This is 1000 times faster than current individual machine learning processors and can be used with high resolution images with 250,000 pixels, large enough for hihg quality facial image recognition applications.
The team used the same hardware to sequentially form an optical convolutional neural network with ten output neurons, achieving successful recognition of handwritten digit images at 88 per cent accuracy.
Optical neural networks offer the promise of dramatically accelerating computing speed using the broad optical bandwidths available. The high performance of the photonic CNN comes from simultaneously interleaving temporal, wavelength and spatial dimensions through an integrated microcomb source. This approach is scalable and trainable to much more complex networks for demanding applications such as autonomous vehicles and real-time video recognition says the team.
"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 who lead the research, published in Nature.
The micro-combs work with hundreds of high-quality infrared lasers on a single chip. They are much faster, smaller, lighter and cheaper than any other optical source.
"This 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," says co-lead author of the study, Dr Xingyuan (Mike) Xu, postdoctoral fellow with the Electrical and Computer Systems Engineering Department at Monash University (above).
"We're currently getting a sneak-peak 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," he said.
The team included other researchers from RMIT University, City University of Hong Kong, Xi'an