Brain on a chip

According to Karlheinz Meier, a physicist at Heidelberg University: "I believe the systems we are going to develop could form part of a new revolution in information technology." Meier is coordinating the EU supported FACETS project, which brings together scientists from 15 institutions in seven countries to build a neural computer. The brain differs from everyday computers in three ways: it consumes very little power; it works well, even if components fail; and it seems to work without any software. The FACETS team is trying to understand why. It's studying neurons to find out exactly how they work, how they connect to each other and how the network can 'learn' to do new things. One part of the team is recording data from neural tissues, while another group is developing mathematical models to describe the behaviour being uncovered. The goal is to use these models to build a 'neural computer' which emulates the brain. The first effort is a network of 300 neurons and 500,000 synapses on a single chip. The team used analogue electronics to represent the neurons and digital electronics to represent communications between them. This 'stage 1' network was designed before the mapping and modelling work. Now, the team is working on a network of 200,000 neurons and 50million synapses that will incorporate the neuroscience discoveries. The team is creating the device on a 200mm silicon wafer and says this approach will make for a more compact device.