The software ecosystem will be different from von Neumann architectures,
designed for sequential operation
Researchers from IBM have achieved breakthrough in developing software architecture to programme silicon chips called SyNAPSE, which is inspired by human brain.
Based on neural architecture, the compact chips will consume low power which will help in developing a new generation of intelligent sensor networks that is similar to the brain’s abilities of perception, action, and cognition.
The system will be different compared to modern computing systems, which were designed decades ago for sequential processing according to a pre-defined programme, based von Neumann architectures.
The new ecosystem is instead based on a scalable, interconnected, configurable network of "neurosynaptic cores", the researchers said.
Each core brings closer memory ("synapses"), processors ("neurons"), and communication ("axons") to execute tasks in an event-driven way.
IBM Research principal investigator and senior manager Dr. Dharmendra S. Modha said architectures and programmes are closely intertwined and a new architecture necessitates a new programming paradigm.
"We are working to create a FORTRAN for synaptic computing chips. While complementing today’s computers, this will bring forth a fundamentally new technological capability in terms of programming and applying emerging learning systems," Modha said.
The company had demonstrated the building block for SyNAPSE chips in August 2011.
IBM along with other collaborators, Cornell University and iniLabs, have recently received $12m funding from Defense Advanced Research Projects Agency to work on the Phase 3 of the Systems of Neuromorphic Adaptive Plastic Scalable Electronics (SyNAPSE) project.
In longer term, the company aims to build a chip system with ten billion neurons and hundred trillion synapses, which will consume just one kilowatt of power and occupying less than two litres of volume.