Scientists have developed a computer chip that attempts to mimic a brain’s functions and contains 5.4 billion transistors helping its artificial intelligence, in an interconnected web, as complex as the brain’s neural network.
The processor, TrueNorth, was described in an article published in the Journal Science on August 7. One of the key features of the chip is that despite having a large number of transistors the chip consumes only 70 milliwatts of power. A normal Intel processor with 1.4 billion transistors consumes 30-140 watts.
The chip was developed by researchers at IBM and attempts to mimic the way a human brain function and recognize patterns. For example, the chip’s “neurons” would be able to tell whether the light is growing brighter changing shape or colour. The chip would then be able to tell whether a man is getting out of the car, or a woman pulling out her purse, taking out a dime and such other activities.
Some of the present computers and robots with artificial intelligence are unable to process these data correctly. Additionally, it has the capacity of storing a tremendous amount of data and experts said that it would excel in calculating an astounding amount of data and put to shame some of the present day supercomputers.
According to IBM researchers, the chip is capable of performing 46 billion operations a second per watt of energy consumed. Although the chip is at its infancy and not available commercially, yet it has generate a keen interest in the tech world.
