Supercomputer with 228,000 billion synaptic operations per second

The DeepSouth supercomputer has an advanced human brain simulation design, expected to start operating in April 2024 .

The DeepSouth supercomputer is the first machine that can simulate the human brain at full scale. (Photo: Western Sydney University)

A group of scientists at Western Sydney University introduced DeepSouth , the first supercomputer with the ability to simulate the human brain at full scale, IFL Science reported on December 19. When operational, DeepSouth will be capable of performing up to 228,000 billion synaptic operations per second. This number is comparable to the activity level of all interconnected nerve cells in the human brain.

The brain is an extremely efficient energy consuming system and to this day, scientists are still trying to recreate that efficiency in computers. Oak Ridge National Laboratory's Frontier supercomputer, among the fastest in the world today, requires 22.7 MW to operate, according to Domenico Vicinanza, associate professor of Intelligent Systems and Science Data at Anglia Ruskin University. However, the human brain can operate at the same speed - a trillion calculations per second - with just 20 W.

DeepSouth plans to help researchers explore computing in a less power-hungry way. This supercomputer has an advanced human brain simulation design, different from the design of traditional electronic computers, which have not changed significantly for decades. DeepSouth is expected to begin operations in April 2024.

DeepSouth's human brain simulation circuit diagram is based on a network of simple processors that can operate in parallel. It mimics the way different neurons in the brain, connected by synapses, can function simultaneously. The system will be scalable and easily reprogrammable from the user interface in the popular Python programming language, meaning researchers can use the technology without needing to understand the hardware.

"This platform will advance our understanding of the brain and develop brain-scale computing applications in many fields such as sensing, biomedicine, robotics, space and AI applications. large scale" , said Professor André van Schaik, director of the International Center for Brain Simulation Systems (ICNS) at Western Sydney University.