[ad_1]
“If you’re trying to understand the brain it’s going to be a tool to do it.”
Brain Stormer
Researchers at Western Sydney University in Australia have collaborated with Intel and Dell technology to build a large computer that aims to simulate neural networks at the scale of the human brain.
They say the computer, named DeepSouth, can simulate neural networks at 228 trillion synaptic operations per second to slow down the mind. , and set equal to the rate at which the human brain completes its tasks.
The project was announced at this week’s NeuroEng Workshop hosted by Western Sydney’s International Center for Neuromorphic Systems (ICNS), a forum for understanding the field of computational neuroscience.
Once operational in April next year, DeepSouth will be able to provide researchers with an unparalleled view of how the user’s brain processes information.
“Progress in our understanding of how the brain uses neurons has been hampered by our inability to model the brain as a network at scale,” said the director of the ICNS and Western Sydney professor André van Schaik in a information.
Sick neurons
Instead of hoping for DeepSouth to become the world’s most powerful conventional supercomputer, the researchers are looking to model the brain’s neuronal network using a “neuromorphic system that simulates biological processes,” in the release. published.
They say the result is a more efficient and much needed supercomputer, built from the ground up to simulate synaptic activity in the human brain.
In simple terms, neuromorphic computing involves the implementation of many operations at the same time but only with the use of small data, which reduces energy.
“The simulation of neural networks on conventional computers using Graphics Processing Units (GPUs) and multicore Central Processing Units (CPUs) is very slow and energy intensive,” van Schaik explained. “Our system will change that.”
The researcher and his team hope to “expand our understanding of the brain and develop brain-scale computing applications in various areas including sensing, biomedical, robotics, space, and large-scale AI applications.” .”
For example, the technology can lead to the development of more intelligent machines or allow AI models to reduce power.
Other researchers who have long been excited about the future of DeepSouth can stop.
“At the end of the day, two types of researchers will be interested in this – either those who study neuroscience or those who want to test new engineering solutions in the AI space,” the Johns Hopkins computer engineering professor Ralph Etienne-Cummings, who was there. participation in the project, tell New Scientist.
“If you’re trying to understand the brain this is the tool to do it,” he added.
More on desktop: China May Have Secretly Built Two Exascale Supercomputers