In what's a significant leap forward in the field of AI, an international team of tech researchers led by the Swinburne University of Technology has developed the world’s most powerful neuromorphic processor for artificial intelligence.
It runs at an astonishing rate of more than ten trillion operations per second (TeraOps/s), which means it will probably process ultra-large-scale data.
Led by Swinburne University’s Professor David Moss, Dr. Xingyuan Xu, and Distinguished Professor Arnan Mitchell from RMIT University, the group enhanced and accelerated computing speed and processing power.
They had been in a position to create an optical neuromorphic processor capable of running over 1,000 times faster than any earlier ones.
The system may also process ultra-large-scale pictures, which is essential for facial recognition characteristics as earlier optical processors have failed in this.
Professor Moss is Director of Swinburne’s Optical Sciences Centre(SOSC), and he was named a top Australian researcher in physics and mathematics within the field of optics and photonics by The Australian.
“This breakthrough was achieved with ‘optical micro-combs,’ as was our world-record web data speed reported in May 2020,” he mentioned.
“We’re currently getting a sneak-peak of ways the processors of the future will look. It’s actually showing us how dramatically we can scale the power of our processors by way of the innovative use of microcombs,” he said.
In accordance with RMIT’s Professor Mitchell, “This technology is applicable to all types of processing and communications — it'll have a big impact. Long term we hope to understand fully integrated systems on a chip, significantly reducing cost and power consumption.”
Professor Damien Hicks supports the international research team and is from Swinburne University and the Walter and Elizabeth Hall Institute.
“Convolutional neural networks have been central to the AI revolution, but existing silicon technology increasingly presents a bottleneck in processing speed and power efficiency,” Professor Hicks mentioned.
“This breakthrough reveals how a brand new optical technology makes such networks faster and better efficiency and is a profound demonstration of the benefits of cross-disciplinary thinking, in having the inspiration and courage to take a thought from one field and using it to solve a basic problem in another,” he said.
It runs at an astonishing rate of more than ten trillion operations per second (TeraOps/s), which means it will probably process ultra-large-scale data.
Led by Swinburne University’s Professor David Moss, Dr. Xingyuan Xu, and Distinguished Professor Arnan Mitchell from RMIT University, the group enhanced and accelerated computing speed and processing power.
They had been in a position to create an optical neuromorphic processor capable of running over 1,000 times faster than any earlier ones.
The system may also process ultra-large-scale pictures, which is essential for facial recognition characteristics as earlier optical processors have failed in this.
Professor Moss is Director of Swinburne’s Optical Sciences Centre(SOSC), and he was named a top Australian researcher in physics and mathematics within the field of optics and photonics by The Australian.
“This breakthrough was achieved with ‘optical micro-combs,’ as was our world-record web data speed reported in May 2020,” he mentioned.
The processor of the Future
Dr. Xu was the co-lead author of the research: “This processor can function a common ultrahigh bandwidth front end for any neuromorphic - optical or digital-based, bringing massive-data machine learning for real-time ultra-high bandwidth data within reach,” Dr. Xu says.“We’re currently getting a sneak-peak of ways the processors of the future will look. It’s actually showing us how dramatically we can scale the power of our processors by way of the innovative use of microcombs,” he said.
In accordance with RMIT’s Professor Mitchell, “This technology is applicable to all types of processing and communications — it'll have a big impact. Long term we hope to understand fully integrated systems on a chip, significantly reducing cost and power consumption.”
Professor Damien Hicks supports the international research team and is from Swinburne University and the Walter and Elizabeth Hall Institute.
“Convolutional neural networks have been central to the AI revolution, but existing silicon technology increasingly presents a bottleneck in processing speed and power efficiency,” Professor Hicks mentioned.
“This breakthrough reveals how a brand new optical technology makes such networks faster and better efficiency and is a profound demonstration of the benefits of cross-disciplinary thinking, in having the inspiration and courage to take a thought from one field and using it to solve a basic problem in another,” he said.
Post A Comment:
0 comments: