The world’s largest supercomputer, designed and built to work the same way as a human brain, has been switched on for the first time. The ‘Spiking Neural Network Architecture’ or ‘SpiNNaker’ machine is capable of completing more than 200 million actions per second.
It has taken 20 years in conception and over 10 years in construction, with the initial build starting in 2006 and has been switched on for the first time. The SpiNNaker machine, which was designed and built in The University of Manchester, can model more biological neurons in real time than any other machine on the planet.
Neuromorphic computing uses large scale computer systems to mimic neuro-biological architectures present in the nervous system. Unlike traditional computers, SpiNNaker doesn’t communicate by sending large amounts of information from one point to another. Instead it mimics the massively parallel communication architecture of the brain, with billions of small amounts of information being sent simultaneously to thousands of different destinations.
SpiNNaker’s creators aim to model a billion biological neurons in real time. One billion neurons is 1% of the human brain, which consists of just under 100 billion neurons, which are interconnected by approximately a quadrillion synapses.
One fundamental use is to help scientists understand how our own brain works. It does so by running large scale real-time simulations which aren’t possible on other machines. SpiNNaker has been used to simulate processing in a range of isolated brain networks.
Neuroscientists can use SpiNNaker to help unlock how the human brain works by running large scale simulations. It can also work as a neural simulator for roboticists to design neural networks into robots so they can walk, talk and move with flexibility and low power.
And has recently been harnessed to control a robot, the SpOmnibot. Which uses the SpiNNaker system to interpret visual information and navigate towards certain objects while ignoring others.