Image
© innovationnewsdaily.com
IBM's research director John E. Kelly has described the next decade in computing at the University of Melbourne, which is where IBM is building and will soon launch a research and development facility

Back in August, IBM announced that it was in the midst of creating neurosynaptic chips to roll in the era of cognitive computing, where computers imitate processes of the human brain. These chips are able to learn, remember, find correlations and create hypotheses through a neurosynaptic core, which consists of an integrated memory (mimicking synapses), communication (mimicking axons) and computation (mimicking neurons).

IBM also showcased a supercomputer earlier this year, named "Watson," who challenged human contestants in a game of "Jeopardy."

Now, IBM's research director John E. Kelly has described the next decade in computing at the University of Melbourne, which is where IBM is building and will soon launch a research and development facility. IBM is looking to bring the era of Watson-like cognitive computing where machines can learn from their environments just like humans.

Kelly described Watson's system, which was $3 million worth of hardware that drew 85 kilowatts of power and was able to understand and search for answers to questions through 500GB of local memory. The contestants, on the other hand, didn't have to study but rather knew the information off the top of their head.

"They have no filing system, the answer is just there and they've done this for so long that they trust the information for the answer will be there when presented with a question," said Kelly. "For those two human beings, the brain consumes only 20W [of power]. The incredible thing is it took 85kW to beat a 40W machine. We need to do bio-inspired computer science; we need to understand how this does what it does. If [computers] can do the type of reasoning we can do with the brain, then we can do some very interesting things."

Kelly estimates that within the next decade, IBM will be able to simulate the human brain's 20 billion neurons and 200,000 billion synapses in exaflop devices, which learn through extracting information from exascale data centers. Exascale supercomputers are the future of IBM's cognitive computing.

"We do not understand how this is wired; we do not understand the fundamentals of how the neurons and synapses are behaving," said Kelly. "All we know is that they're not ones and zeros -- they're multi-state devices, and they learn over time and their behavior changes based on what they have experienced."

Kelly mentioned that exascale supercomputers could be used in several applications, such as healthcare. In fact, IBM already reached an agreement with WellPoint, a U.S. health insurance company, to create Watson-like technology that uses WellPoint's historical data of treatments and results. This would make statistical decisions instead of skewed decisions by doctors.