IBM Scientists Advance Toward Brain-Like Computing, Develop ‘Digital Mouse Brain’
IBM scientists are advancing toward “neuromorphic” computing – digital systems that process information like the brain – and launching a complete ecosystem for brain-like computing, with important near-term applications and visionary long-term prospects.
“For decades, computer scientists have been pursuing two elusive goals in parallel: engineering energy-efficient computers modeled on the human brain and designing smart computing systems that learn on their own – like humans do – and are not programmed like today’s computers,” said Dharmendra S. Modha, IBM Fellow and Chief Scientist for brain-inspired computing.
Both goals are now within reach.
A Digital Brain and a Digital Mind
With the SyNAPSE Project, initially funded by the U.S. Defense Advanced Research Projects Agency (DARPA) with $58 million, IBM Research, in collaboration with Cornell University, developed the TrueNorth neuromorphic chip, able to process large amounts of data efficiently with brain-inspired computing. The research leading to the TrueNorth chip is described in a 2014 paper titled “A million spiking-neuron integrated circuit with a scalable communication network and interface,” published in Science.
Now IBM launched an ecosystem for brain-inspired computing with a TrueNorth Boot Camp for academic and government researchers.
“With 1 million neurons, 256 million synapses, 4096 cores interconnected via a network-on-chip at less than 100mW of power consumption, TrueNorth is literally a supercomputer the size of a postage stamp consuming the power of a hearing aid battery,” said Modha. “The chip has an entirely novel parallel, distributed, modular, scalable, fault-tolerant, event-driven architecture.
Modha added that the brain-inspired computing ecosystem includes development boards, a simulator, a programming language, an integrated programming environment, a library of algorithms and applications, firmware, deep learning tools, a teaching curriculum, and cloud enablement.
Future smartphones could include high-performance, low-power neuromorphic chips that could permit data and computing intensive applications beyond the possibilities of today’s smartphones, such as face recognition.
But the ultimate goals of the project are nothing short of amazing: “The best possible outcome is to map the entirety of existing cache of neural network algorithms and applications to this energy-efficient substrate,” said Modha. “And, to invent entirely new algorithms that were hereto before impossible to imagine.”
Then, to compose these algorithms to exhibit mind-like behavior on top of the brain-like substrate.
In other words, the researchers want to build a digital brain and a digital mind.
TrueNorth is not yet a digital brain, but it is a step toward a digital brain,” notes Wired. And with IBM’s boot camp, the project is accelerating. IBM has built the digital equivalent of a rodent brain with 48 TrueNorth neuromorphic chips. Modha said that the system has 48 million artificial nerve cells, roughly the number of neurons packed into the head of a rodent. He added:
You’re looking at a small rodent.
Images from IBM Research.