How Much Power Does It Take To Simulate The Human Brain?

Kwabena Boahen , a calculator scientist at Stanford University , believes that it would require 10 megawatt to power a central processing unit as smart as the human brain . His new “ Neurogrid ” supercomputer might be able-bodied to do it on only 20 watts .

To put that in perspective , 10 megawatts is the kind of energy a little hydroelectric plant produces—20 watts is only enough juice to power up a weak light bulb . Amazingly , your physical encephalon run on this minuscule amount of magnate , and it ’s not very efficient . However , embracing this inefficiency could be the keystone to creating computers that mimic the human brain .

It sound cockamamy , but it is true . scientist have set up that the brain ’s 100 billion neuron are surprisingly undependable . Their synapsis fail to give notice 30 percent to 90 percent of the sentence . Yet somehow the brainpower works . Some scientists even see neural randomness as the headstone to human creativity . Boahen and a low chemical group of scientists around the world desire to simulate the brain ’s noisy calculations and spawn a raw era of vigor - effective , intelligent computing . Neurogrid is the trial to see if this approach can succeed .

Family Residence Damage Tornado Stlois

Most innovative supercomputer are the sizing of a icebox and consume $ 100,000 to $ 1 million of electricity per twelvemonth . Boahen ’s Neurogrid will tally in a briefcase , run on the equivalent weight of a few 500 batteries , and yet , if all belong well , come close to keeping up with these giant .

So far Boahen has negociate to squeeze a million neurons onto his unexampled supercomputer compare to only 45,000 last year . By 2011 , he hopes to have 64 million up and go , bring the labor to the eq of a computer mouse ’s brain .

Ditching reliability and efficiency in favour of organized bedlam flies in the face of everything that an engineer holds devout , but the approach does make sensory faculty — and decoct the power consumption is the key to upholding Moore ’s police . But how will this development change our sensing of what an artificially intelligent golem might become ? Instead of some cold , coherent machine that can recall for itself , we might end up with robots that are just as stupid and blemished as we are . In other parole . it could be a golem on that instalment of next copper running through the shrub with no shirt on after trying to rob a contraption computer storage with a credit card lightsaber . suppose about it . [ Discover MagviaPopSci ]

Last Of Us 7 Ellie Crash