A team of international scientists has designed cutting-edge computer chips that could support future technologies that involve machine learning and artificial intelligence.
A team of international scientists has designed cutting-edge computer chips that could support future technologies that involve machine learning and artificial intelligence.
UNSW Canberra Associate Professor Andrey Miroshnichenko explained that new design of optical chips enables light to experience multiple dimensions.
“We are living in a 3D physical world, which puts limitations on what can be studied experimentally,” Associate Professor Miroshnichenko said.
“Here we propose a robust approach, which allows us to study and explore higher-dimensional effects in tabletop experiments.”
Associate Professor Miroshnichenko worked with researchers from the Australian National University (ANU), the University of Rostock in Germany and the University of Central Florida.
Professor Alexander Szameit from the University of Rostock led the experimental work, including the fabrication of optical circuits.
“Making use of higher dimensions on optical chips could support a variety of future technologies that involve machine learning and performing complex tasks autonomously,” Professor Szameit said.
ANU Professor Andrey Sukhorukov said light holds great promise for information processing in the next industrial revolution.
“Light can evolve in up to seven dimensions on our specially designed circuits, which is mind boggling when you realise that the space around us is three-dimensional,” Professor Sukhorukov said.
Associate Professor Miroshnichenko said one of the immediate applications for the research is the design of novel photonic integrated circuits for unprecedented signals processing.
“The developed approach is much more general and has an ability to enable next-generation artificial intelligence technologies based on compact and drastically simplified neuromorphic chips,” he said.
Synthesizing multi-dimensional excitation dynamics and localization transition in one-dimensional lattices was published by Nature Photonics: https://doi.org/10.1038/s41566-019-0562-8