The computer, called DeepSouth, has the capacity to emulate neural networks at a speed of 228 trillion synaptic operations per second; This corresponds to the estimated speed at which the human brain performs operations.

Detail

The project was announced this week at the NeuroEng Workshop hosted by the International Center for Neuromorphic Systems (ICNS) in Western Sydney, a forum for computational neuroscience professionals.

When it becomes operational next April, DeepSouth will be able to offer researchers a unique perspective on how the human brain processes information.

Progress in our understanding of how the brain processes neurons has been hampered by our inability to model brain-like networks at scale.
ICNS Director and Western Sydney Professor Andre van Schaik said in a statement.

Instead of building DeepSouth into the world’s most powerful conventional supercomputer, researchers aim to mimic the brain’s neural network using a “neuromorphic system that mimics biological processes,” according to a press release.

The result, scientists say, will be a more efficient and less energy-intensive supercomputer built from the ground up to simulate synaptic activity in the human brain.

Simply put, neuromorphic computing involves performing large numbers of operations simultaneously while moving very little data and making them consume much less energy.

The team hopes to “improve our understanding of the brain and develop brain-scale computing applications in a variety of fields, including sensing, biomedicine, robotics, space, and large-scale artificial intelligence applications.” For example, this technology could lead to the development of advanced smart devices or allow AI models to consume less energy.