A computer chip designed to mimic the performance of the human brain? Yes. The chip is developed by IBM together with Cornell Tech and could prove a big step forward in the future of computing power. The chip is called SyNAPSE, which stands for Systems of Neuromorphic Adaptive Plastic Scalable Electronics.
So whats the human brain part about?
This chip is capable of 1 million programmable neurons, 256 million programmable synapses and 46 billion synaptic operations per second, per watt. IBM has also tethered 16 of these chips together in four four-by-four arrays, which collectively offer the equivalent of 16 million neurons and 4 billion synapses, showing that the design can be easily scaled up for larger implementations. Unlike traditional chips, who follow the lines of the von Neumann-architecture meaning that they process information step by step, this chip is able to be active on different levels and actions.
IBM calls this cognitive computing because of a dynamic that attempts to mimic the interactions of neurons and synapses in biological brains. It offers more of a organic approach to problem solving, based on hypotheses, past experiences and trail and error. Similar to a human brain.
The chip is about the size of a postage stamp and can run on a small amount of energy, about equivalent to what it takes to power a hearing aid. Such a chip could act as a low-power sensor for a range of embedded and portable devices. Future applications according to the research could include solar-powered, leaf-shaped sensor modules that could send out environmental and forest fire alerts.
It is most of all a architectural breakthrough that could prove to be crucial for next-generation cloud and big-data processing. It could eventually mean that we end up with much faster, cooler and more compact ways of (cognitive) computing. If you’d like to know more, the researchers describe the chip in this week’s issue of the Science journal.