The world ‘s largest neural computer, designed and built to work the same way a human brain works, took its 1 million core processor; and it was first activated on November 2nd.
The “SpyNNaker (Spiking Neural Network Architecture)” supercomputer is able to perform over 200 million million operations per second with each of its chips having 100 million semiconductors. In order to reach this point, it was necessary to finance 15 million pounds, 20 years to conceive the idea and over 10 years to build the supercomputer; with the initial work to begin in 2006. The project was originally financed by the “EPSRC” and is now supported by the “European Human Brain Project” .
The “SpiNNaker” designed and built by the “School of Computer Sciences” of “The University of Manchester”, and can simulate the function of more biological neurons in real time from any other machine in the world. The biological neuron is basically brain cells in the nervous system which communicate primarily by emitting “surges” of electrochemical energy. Neuromorphous computers are large computer systems with electronic circuits that mimic these “bouts” .
The difference between “SpiNNaker” and “traditional” computers is that it does not communicate by sending large volumes of information from point A to point B through a conventional network, but rather imitates the parallel communication architecture of the brain, including simultaneous sending billions of small packets of information simultaneously to thousands of different destinations.
“SpiNNaker is revising the way conventional computers work. In fact, we built a machine that works more like a brain than a traditional computer,
said Steve Fourmper, professor of “Computer Science”, who had the original idea.
“The ultimate goal for the program has always been one million cores on a computer for real-time brain model applications, and now we’ve done it, which is fantastic.”
The creators of the computer in the long run aim to simulate the operation of up to a billion of biological neurons in real time.
Indicatively, to figure out the sizes, a mouse brain consists of 100 million neurons, and the human brain is 1,000 times larger. One billion neurons are 1% of the human brain, consisting of less than 100 billion brain cells (neurons).
Regarding the computer uses one of these will be a better understanding of how the human brain works; and recently the power utilized for controlling a robot, the “SpOmnibot”, using the supercomputer system for interpreting visual information in real time to move around in places directed to specific objects and avoid obstacles.
(Συνολικές Επισκέψεις: / Total Visits: 26)
(Σημερινές Επισκέψεις: / Today's Visits: 1)
Σας αρέσει το άρθρο; / Do you like this post?