For quite some time scientists have understood that the human brain is the most efficient and powerful computer on planet Earth. Consuming some 40,000 times less energy than traditional computers the brain is the ideal machine for blasting through complex simulations or processing large data sets (like reality). With that potential in mind, scientists have spent decades working on ways to create a chip that’s modeled after the brain, hoping to harness some of its incredible power. According to Stanford University, some of its researchers have done just that.
Named the Neurogrid, the iPad sized computation device consists of 16 special “neurocores”, each of which simulates the lightning quick synapsis that facilitate communication in the brain. Loosely networked into groups, a unit of neurocores can process information nine-thousand times faster than a traditional PC while performing 100,000 times more efficiently.
Though Stanford’s Neurogrid is a powerful machine it still pales in comparison to the wet wear enfolded in our heads. According to Kwabena Boahen, associate professor of bioengineering, "The human brain, with 80,000 times more neurons than Neurogrid, consumes only three times as much power," Boahen continues, "Achieving this level of energy efficiency while offering greater configurability and scale is the ultimate challenge … engineers face."
Beyond problems of efficiency and scale, the Neurogrid faces the challenge of cost. At the moment the lab’s prototype costs $40,000 to build. Although prohibitively expensive at the moment, Boahen and his team believe they can reduce the cost of their machine 100-fold, producing powerful machines for around $400 each.
Although Stanford’s Neurogrid project falls short of achieving the hyper efficiency of the human brain its abilities are nevertheless remarkable. If Boahen’s dream of an affordable Neurogrid are realized it could be a step towards technology that forms the backbone upon which advanced robotics, sensory-capable prosthetics and artificial intelligence are built; and given the complexity of the problems associated with those fields any progress is incredible news.
Image Courtesy of Stanford