IBM Developing Next-Gen Computer That Mimics Human Brain

That may sound like a scene from a low-budget, science-fiction movie, the one where a giant brain in a glass case is the central processing system for a futuristic society. But the IBM project, which has won $4.9 million in funding from the Pentagon's Defense Advanced Research Projects Agency (DARPA), is very real.

IBM announced Thursday that its "Cognitive Computing via Synaptronics and Supercomputing" (C2S2) project will attempt to design computing systems that "simulate and emulate the brain's abilities for sensation, perception, action, interaction and cognition," while rivaling the brain's compact size and low power consumption.

The idea of building computers that mimic the human brain isn't new: Development of artificial intelligence technology is almost as old as computing itself. While aspects of AI have found their way into expert systems, pattern recognition applications and other technologies, AI has never quite lived up to its early hype.

IBM said the volume of digital data that computer systems have to process today is growing exponentially, while the ability of today's information technology to monitor, analyze and react to that information is lagging. Cognitive computing systems could "integrate and analyze vast amounts of data from many sources in the blink of an eye," IBM said in a statement.

The C2S2 project will include research over the next nine months in the areas of synaptronics, material science, neuromorphic circuitry, supercomputing simulations and virtual environments, IBM said. Initially, the research will focus on developing nanoscale synapse-like devices that will link many computers around the world into a single "cognitive computer" in much the same way synapses in the human brain connect neurons or nerve cells.

IBM said C2S2's long-term goal is to develop "low-power, compact cognitive computers that approach mammalian-scale intelligence." IBM said it recently assembled a system equivalent to the brain of a small mammal using cognitive computing algorithms and its BlueGene supercomputer.

Working with IBM on the project are noted researchers at Stanford University, the University of Wisconsin-Madison, Cornell University, Columbia University Medical Center and the University of California-Merced.

DARPA awarded the funding as part of the first phase of its own Systems of Neuromorphic Adaptive Plastic Scalable Electronics (SyNAPSE) initiative.