Date Published: 7 July 2011
Chips based on ARM processor technology to mimic human brain functions
Scientists at Manchester University (England) are working towards producing a high-performance computer which they intend to create working models of human brain functions. Microprocessor chips (such as are used in computers and other electronic devices) will be linked together to simulate the highly-complex workings of the brain, whose functionality derives from networks of billions of interacting, highly-connected neurons.
The "chips" used for this project are based on ARM processor technology. They are critical to the success of this work, were delivered last month, and have passed all functionality tests so far. These chips will form the system architecture for a massive computer, called SpiNNaker (Spiking Neural Network architecture), which it is hoped will map out the brain's individual functions.
SpiNNaker could become a vital tool for neuroscientists, psychologists and medical doctors to help these professionals understand complex brain injuries, diseases and conditions, and identify the most effective therapies.
Even though there will be up to one million ARM processors the same technology as used in most of the world's mobile phones in the final SpiNNaker machine, this is only expected to enable computer scientists to recreate models of only up to 1% of the human brain.
An important challenge is developing and understanding the information processing of the brain and the extremely high connectivity of the brain cells. There are 100 billion neurons with 1,000 trillion connections in the human brain. In this system, the neurons emit spikes which are relayed as tiny electrical signals. Each impulse is modelled in SpiNNaker as a 'packet' of data, a very-much scaled down version of the way the internet carries information. This packet is then sent to all connected neurons. Neurons are represented by small simple equations which are solved in real-time by software running on the ARM processors.
The electronic connections in SpiNNaker convey these spikes much quicker than the biological connections in the brain, hence SpiNNaker can transmit spikes as effectively and quickly with many fewer connections.
About the Microchips
Test chips were first developed in 2009. The latest full chips, designed in Manchester and manufactured in Taiwan, have now been delivered and will allow the scientists to develop much larger systems which can model the brain. Each chip contains 18 ARM processors. The bespoke microchips are integrated in a single 19mm square package with a second microchip that provides substantial memory using 3D System-in-Package technology from Unisem Europe Ltd, who have packaging facilities in south Wales. This package delivers the computing power of a PC in a tiny space and for around one watt of electrical power.
Professor Furber, from the School of Computer Science, said: "We hope the machine will be able to model important functions of the human brain and help us gain key insights into their principles of operation.
" Developing and understanding the information processing in the brain is the key. We are actively engaging with neuroscientists and psychologists, both here at the University and elsewhere. This could ultimately be of great help for patients, for example, who have presented with reading problems caused by strokes or similar brain injuries. Psychologists have already developed neural networks on which they can reproduce the clinical pathologies. They then use these networks to test alternative therapies, to identify which is most effective in treating the patient's symptoms. At present they are limited in the fidelity they can achieve with these networks by the available computer power, but we hope that SpiNNaker will raise that bar a lot higher. We don't know how the brain works as an information-processing system, and we do need to find out. We hope that our machine will enable significant progress towards achieving this understanding."
University, England (UK)