The K is Coming
Earlier this month, the Top500, the project which measures and ranks the world’s fastest supercomputers, unveiled the latest instalment of their twice-yearly list. It had a new number one – Japan’s K computer.
Of course, an even faster computer is in itself very exciting, this is especially exciting because the project, pronounced kei, aims to be the first computer to reach ten petaflops per second when it becomes fully operational in November 2012.
Ten petaflops is a key number because, despite there being much discussion of its accuracy, ten peraflops is the number put forward by Kurzweil for the upper boundary of estimates on the processing power of the human brain.
That means that, once the K is fully operation, for the first time we will have a computer more powerful than the human brain.
That’s pretty exciting!
Of course, it could be entirely inaccurate. Some think the brain is capable of 38 petaflops per second, or even higher – other estimates have suggested 100, or even 1000 petaflops.
But considering the exponential growth of computer power, even if that is true, that doesn’t actually delay the arrival of such a computer that much time.
Consider Cray’s new XK6. It is aiming to hit 50 petaflops (http://www.informationweek.com/news/hardware/supercomputers/229700091) and they say it will be ready later this year! Of course, it hasn’t been delivered yet, but presuming it does, this represents a significant step forward in the chase to beat the brain.
Even if you assume that the brain does, in fact, operate at 1,000 petaflops per second, 100 times faster than Kurzweil suggested, the release of the XK6 this year means that within seven years, 2018, we will still achieve a computer faster than the human brain. Soon enough that I very much hope my grandparents will still be around to see it.
Earlier this month, the Top500, the project which measures and ranks the world’s fastest supercomputers, unveiled the latest instalment of their twice-yearly list. It had a new number one – Japan’s K computer.
Of course, an even faster computer is in itself very exciting, this is especially exciting because the project, pronounced kei, aims to be the first computer to reach ten petaflops per second when it becomes fully operational in November 2012.
Ten petaflops is a key number because, despite there being much discussion of its accuracy, ten peraflops is the number put forward by Kurzweil for the upper boundary of estimates on the processing power of the human brain.
That means that, once the K is fully operation, for the first time we will have a computer more powerful than the human brain.
That’s pretty exciting!
Of course, it could be entirely inaccurate. Some think the brain is capable of 38 petaflops per second, or even higher – other estimates have suggested 100, or even 1000 petaflops.
But considering the exponential growth of computer power, even if that is true, that doesn’t actually delay the arrival of such a computer that much time.
Consider Cray’s new XK6. It is aiming to hit 50 petaflops (http://www.informationweek.com/news/hardware/supercomputers/229700091) and they say it will be ready later this year! Of course, it hasn’t been delivered yet, but presuming it does, this represents a significant step forward in the chase to beat the brain.
Even if you assume that the brain does, in fact, operate at 1,000 petaflops per second, 100 times faster than Kurzweil suggested, the release of the XK6 this year means that within seven years, 2018, we will still achieve a computer faster than the human brain. Soon enough that I very much hope my grandparents will still be around to see it.