By 2018, supercomputers could operate 100 times faster than the human brain
December 2, 2010 by Amara D. Angelica
The breakthrough (see Breakthrough Chip Technology Lights the Path to Exascale Computing) announced Wednesday by IBM researchers has been long sought: a way to use pulses of light in waveguides instead of electrons in wires for chip connections. Electrons generate heat, which limits has fast chips can work and requires a lot of power for cooling. Light has no such limitations.
The new technology could accelerate the performance of supercomputers a thousand times, taking us from the current 2.6 petaflops (1015 or quadrillion operations per second) Chinese Tianhe-1A supercomputer to an exaflop (1018 or a quintillion operations per second) supercomputer in just 8 years (“flops” stands for “floating point operations per second”).
And that means a supercomputer that runs 100 times faster than a human brain operating at peak capacity, currently estimated to be around 1016 operations per second by Ray Kurzweil and others.
That doesn’t mean a machine would be smarter than a human; that would require software. Really smart software. More on that next week.
The wild ride ahead. Blue Waters is a 10 petaflops IBM supercomputer scheduled for 2011. HPCS = "high-performance computer systems" (Credit: IBM)
Comments (3)
by gmn17
100,000 processors @ 10 TFlop each = 1ExaFlop
1024 cores per processor * 100,000 = 100,000,000 cores
by omelv44
Sourse Wikipadia: “Daniel Lyons, writing in Newsweek, criticized Kurzweil for some of his predictions which turned out to be wrong; such as the economy continuing to boom from the 1998 dot-com through 2009, a US company having a market capitalization of more than $1 trillion, a supercomputer achieving 20 petaflops, speech recognition being in widespread use and cars that would drive themselves using sensors installed in highways; all by 2009.”
You are too much optimistic Mr. Kurzweil. And as we can see, you turned out to be wrong! :) ;)
by omelv44
Now I am sure that I was right when I said that Kurzweil is a secret PR agent of the IT companies. To make his predictions come true, the IT companies may simply help him by sharing the insider information to him. The most modern technologies reach a usual consumer with a time lag which can be very very large. E.g. the concept of the Internet appeared in 50-s but it became available to anyone not so many years ago. Moreover, when some new technology appears, the inventors may find to it some better use than the sail to the general public. And at this time they may say such guys as Kurzweil to predict that this new product will appear in some year. And voila, it appears because the inventors don’t need it anymore so they are willing to sell it. And this is how Kurzweil gains his reputation of an accurate predictor of the future! He mixes the insider information voluntarily provided to him by the IT firms with the bullshit aiming to prove the the Silicon valley is capable of creating a magic computer whose appearance will turn out with the technological singularity. And they do it for 1 reason: it augment the stock prices of the IT companies. Do you think that Kurzweil predicted the burst of the dot-cot bubble? No! According to his predictions the economy should be continuing to boom from the 1998 dot-com through 2009, a US company having a market capitalization of more than $1 trillion. What the fuck! It burst and many investor lose all their money because of his bullshit! Why didn’t he warn us on the impending threats? Because it is not his job! His job is to promise the moon. Because the IT companies want it. He also didn’t predict the crisis. He made such mistakes in his predictions that are outside of the insider information, that it is unpardonable to him.
This truth must be known! Please spread it to all your friends and people who may know about Ray Kurzweil.