NeurosynapticComputingChip
Latest
IBM's new supercomputing chip mimics the human brain with very little power
A lot has changed in the three years since IBM first unveiled a prototype of its human brain-inspired SyNAPSE (Systems of Neuromorphic Adaptive Plastic Scalable Electronics) chip. That single-core prototype has now been significantly scaled up, leading to a new, production-ready SyNAPSE chip that blows past its predecessor with 1 million neurons, 256 million synapses and 4,096 neurosynaptic cores, all the while only requiring 70mW of power. Though the numbers are impressive, it's what they translate to that holds even greater prominence: the ability for devices to process various sensory data in parallel just like the human brain, by merging memory and computing.
IBM's cognitive computing chip functions like a human brain, heralds our demise (video)
After having created a supercomputer capable of hanging with Jeopardy's finest, IBM has now taken another step toward human-like artificial intelligence, with an experimental chip designed to function like a real brain. Developed as part of a DARPA project called SyNAPSE (Systems of Neuromorphic Adaptive Plastic Scalable Electronics), IBM's so-called "neurosynaptic computing chip" features a silicon core capable of digitally replicating the brain's neurons, synapses and axons. To achieve this, researchers took a dramatic departure from the conventional von Neumann computer architecture, which links internal memory and a processor with a single data channel. This structure allows for data to be transmitted at high, but limited rates, and isn't especially power efficient -- especially for more sophisticated, scaled-up systems. Instead, IBM integrated memory directly within its processors, wedding hardware with software in a design that more closely resembles the brain's cognitive structure. This severely limits data transfer speeds, but allows the system to execute multiple processes in parallel (much like humans do), while minimizing power usage. IBM's two prototypes have already demonstrated the ability to navigate, recognize patterns and classify objects, though the long-term goal is to create a smaller, low-power chip that can analyze more complex data and, yes, learn. Scurry past the break for some videos from IBM's researchers, along with the full press release.
