Laser neurons revolutionise AI with unprecedented speed
Scientists from the University of Hong Kong have developed a laser neuron that operates a billion times faster than the biological nerve cells in our brains. This innovative chip will accelerate the performance of artificial intelligence further.
Researchers from China have created an artificial neuron based on laser technology that mimics the functions of biological neurons in the brain but operates at much higher speeds. It emulates the dynamics and information processing methods characteristic of living neurons, and its speed surpasses natural cells by an astounding billion times.
The new neuron processed data from 100 million heartbeats in one second
As the team of experts explained, living organisms possess two types of neurons: those with gradual responses and those with impulsive responses. The solution discussed focuses on gradual neurons, which process signals more precisely and thus bypass the limitations of current photonic impulse structures.
Team leader Chaoran Huang emphasises that the new laser neuron features exceptional dynamics and rapid data processing, making it ideal for pattern recognition or sequence prediction tasks. The new neural system's unique properties enable it to operate like a miniature neural network, executing advanced tasks even as a single unit.
In tests, the system demonstrated the ability to process vast amounts of data in a fraction of a second. Thus far, laser-based neurons have processed data from 100 million heartbeats and nearly 35 million digital images in just one second.
Scientists: the potential of artificial neurons can be further increased
Professor Huang, a co-author of the publication, points out that using cascading laser neurons will offer even greater potential.
"In this work, we used a single laser graded neuron, but we believe that cascading multiple laser graded neurons will further unlock their potential, just as the brain has billions of neurons working together in networks," explains Professor Huang.