Naked Science Forum
General Science => General Science => Topic started by: savyinvestor on 13/10/2006 06:08:00
-
Does anyone know how fast a quantum computer could run?
I know that in June of 2006 some super computer achieved 1 petaflop,(1,000 tera flops, or 1,000 trillion calculations/sec).
The Japanese are working on one that is supposed to be 73 times faster than the IBM Blue Gene, which would comes to just under 10 peta flops.
With the limit of transistor based integrated circuitry expected in 2010 or so, I fear that mankind will never achieve the all important computation speed of 1 trillion exoflops, (1 trillion *1 million* 1 trillion calculations/sec).
At this speed computers will be able to accurately decode all genetic information for any organism on the planet within 24 hours, including what each gene does and why/how.
Such a machine could be used to determine the key to stopping aging and achieving immortality, (needed if we are to collonize space.)
So does anyone know where one can find the potential speed of a quantum computer, or perhaps a DNA computer?
All replies are greatly appreciated.
-
Quantum computing is not quite there yet. Theoretically it's the way to go; for those not in the know, whilst a "normal" computer works in binary 1's and 0's, or black and white to put it another way, quantum information is almost like the grey scale between the two, so the amount of information that can be stored is huge.
The slight snag at the moment is that although we know how to manipulate this information and how to utilise the technique we are nowhere near doing so at a practical level. There was a paper in a recent edition of Nature in which Swiss scientists had succeeded for the first time in teleporting quantum information between two different forms of matter - light and a solid (some atoms). This is one way that a quantum network could move information around in future, but it's very early days.
At the moment, the computer industry is facing "red wall" in about the next 5 years. This is the theoretical maximum computing power achieveable using present silicon technology. The basis of it is that as you pack more and more components onto a chip at greater densities and smaller sizes, eventually the components become so close to each other that they begin to interfere with each other.
The solution may lie in using nano-assemblers. In other words structures with a defined shape which can arrange tiny clusters of atoms very precisely in space to make components. And because they are so tiny the density can be very high and the time taken for then to communicate with each other is therefore very low (because they are so close together). Scientists are now exploring the possibility of using DNA, proteins, or even chlorophyll (the chemical that makes plants green) for this purpose.
I recently interviewed nanotechnologist Professor Donald Fitzmaurice (http://www.thenakedscientists.com/HTML/shows/2006.02.26.htm), who works on this technology.
Chris
"I never forget a face, but in your case I'll make an exception"
- Groucho Marx
-
A quantum computer isn't necessarily very fast, but even if it could only do one calculation in a second there are some problems that it would be far faster than a petaflop computer because it can do operations that a conventional computer can't.
The sort of problem that a conventional computer is very slow at which a quantum computer would be very good at are the ones where you are trying to find one out of billions of billions of billions of combinations which produces an answer. A conventional computer has to go through all the possibilities one by one, the quantum computer can in some sense try them all out at once and can therefore do the calculation in far fewer steps.
The classic problem they would be good at, is finding the factors of very large numbers, which coincidently is what you would need to do to crack the encryption you use when you buy things on the internet... for general use I think conventional computers are going to be around for a while.
-
Not sure that simply measuring the speed of computing, at least insofar as one looks at operations per second, really has any meaning.
The thing you want to look at a cost, power, and size. You can always create more operations per second simply by tying together lots of smaller computers, but this uses space, generates lots of heat, requires lots of energy, and is expensive, and finally only works if the problem can be heavily parallelised.
Quantum computing is compact, and low energy; but the problem is that at present the supporting infrastructure (not least the refrigeration) is large, expensive, and energy consuming. The question has to be not only what one single logic device can do, but what the entire system is capable of, including all the ancillary hardware needed to make the logic work.
As Dave has mentioned, there may well be some particular problems that quantum computers can specialise in, just as vector processors can be used in some scenarios, and not in others; but this is very different from speeding up general purpose computing.
George
-
So a quantum computer can calculate different scenarios simultaneously, thus requireing "less steps" as some one said.
Wouldn't this still allow a speed nominclature of tera, peta, exa flops?
If a super computer checks scenarios 1 by 1, but at 1 trillion times in a second, which I will call a cycle, and a quantum computer checks 10 trillion scenarios simultaneously, in a 1 second cycle, then could I not describe the quantum computer as operating at 10 Teraflops?
And since there is an infinite continum between 0-1 shouldn't there be no limit on potential speed?
Shouldn't 1 quintillion exaflops (1 million trillion million trillion calculations/sec) be possible at some point?
-
quote:
Originally posted by savyinvestor
So a quantum computer can calculate different scenarios simultaneously, thus requireing "less steps" as some one said.
Wouldn't this still allow a speed nominclature of tera, peta, exa flops?
If one simply wants to measure things based upon simple numbers, then yes, it will allow such nomenclature. The problem is whether you would hire a grand prix racing car to move furniture around – speed alone does not tell you all you need to know.
quote:
And since there is an infinite continum between 0-1 shouldn't there be no limit on potential speed?
I think you may find the Heisenberg's uncertainty principle pops its head up here somewhere.
George
-
I've read quite a lot of literature on quantum computing & I have to say I find the whole subject mind-boggling - not to say confusing. I think the uncertainty principle applies to my understanding of quantum, full stop!
Brand new forum at
http://beaverland.forumup.us
More than just science
-
quote:
Originally posted by DoctorBeaver
I've read quite a lot of literature on quantum computing & I have to say I find the whole subject mind-boggling - not to say confusing. I think the uncertainty principle applies to my understanding of quantum, full stop!
I understand you!