The European Commission will fund for the next 4 years the large scale project FACETS (acronym for "Fast Analog Computing with Emergent Transient States") in the framework of the FET ("Future Emergent Technologies") initiative.
Coordinated by the Kirchhoff-Institute for Physics at the Ruprecht-Karls-Universität Heidelberg 13 leading European research groups working in the fields of neuroscience, computer science, electrical engineering and physics, from Austria, France, Germany, Hungary, Sweden, Switzerland and the UK have joined forces.
The idea of the project is to extract basic principles of information processing in the human brain using biological data, to simulate these principles on state-of-the-art computers and finally to implement them on custom-made analog microchips. The microchips will be used to set up massively parallel networks following the biological principles initially extracted from the neuroscience data. The project title contains already a hint to an important physics concept relevant for biological information processing : The concept of Time. The precise temporal arrangement of action potentials (spikes) emitted by individual neural cells plays a major role for the mechanisms of self-organisation which distinguish such set-ups from the conventional neural networks known from computer science.
The FACETS project bundles the skills of experimental and theoretical neuroscientists, the numerical ability of multi-node computing clusters and the circuit technology of modern microelectronics on the 100 nanometer scale. To perform simulations of neural microcircuits the number 8 high performance computer BlueBrain at the EPFL Lausanne (Brain-Mind-Institute) is available at one of the consortium partners. The Heidelberg physics group contributes experience in the design and construction of complex analog circuits emulating the biological example.
As a along-term goal the project work aims towards the development of novel systems for information processing working according to principles radically different from those implemented in contemporary IT systems. Such new systems will exhibits features like full scalability, ability to adapt and self-organise, very low power consumption and fault tolerance. These features may eventually open up new and interesting applications.
More information about FACETS and media material available for download can be found on the project website :
The material includes a schematic drawing of a neural processing unit designed for he project.
The website of the Heidelberg group can be found here:
Contact and detailed information :
Prof. Dr. Karlheinz Meier
FACETS Project Coordinator
Kirchhoff-Institut für Physik
Im Neuenheimer Feld 227
Phone.: 00 49 6221 54 9830
Fax : 00 49 6221 54 9839
e-Mail : firstname.lastname@example.org
Dr. Michael Schwarz
Press Officer of the University of Heidelberg