Supercomputer with 228,000 billion synaptic operations per second

The DeepSouth supercomputer, designed for advanced human brain simulation, is set to commence operations in April 2024.

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

A team of scientists at Western Sydney University has unveiled DeepSouth, the first supercomputer capable of fully simulating the human brain, according to IFL Science on December 19. Once operational, DeepSouth will be able to perform up to 228,000 billion synaptic operations per second, mirroring the activity level of all interconnected nerve cells in the human brain.

The brain is an extraordinarily efficient energy consumer, and replicating this efficiency in computers remains a challenge for scientists. The Frontier supercomputer at Oak Ridge National Laboratory, one of the fastest globally, requires 22.7 MW to operate, notes Domenico Vicinanza, an associate professor of Intelligent Systems and Science Data at Anglia Ruskin University. In contrast, the human brain can achieve the same computational speed – a trillion calculations per second – with just 20 W.

DeepSouth aims to assist researchers in exploring computing methods that consume less power. Unlike traditional electronic computers, whose designs have remained largely unchanged for decades, DeepSouth features an advanced human brain simulation architecture. It is anticipated to begin operations in April 2024.

The design of DeepSouth’s human brain simulation circuit is based on a network of simple processors that can function in parallel, emulating how different neurons in the brain, connected by synapses, operate simultaneously. The system will be scalable and easily reprogrammable through a user interface utilizing the widely-used Python programming language, allowing researchers to employ the technology without needing in-depth hardware knowledge.

“This platform will enhance our understanding of the brain and foster the development of brain-scale computing applications across various fields such as sensing, biomedicine, robotics, space, and large-scale AI applications,” stated Professor André van Schaik, director of the International Center for Brain Simulation Systems (ICNS) at Western Sydney University.