Go to bottom

do you believe in quantum computing ?

category: general [glöplog]
I don't. Even if it exists. I don't think you can make a general purpose quantum processor.
added on the 2007-11-09 20:31:41 by xernobyl xernobyl
some company already sells a quantum computer with a few bits
of "processing power" or how it is called in that context.
it's either a german or austrian company... can't remember now.

read about it in c't, a quite popular german computer-mag.
added on the 2007-11-09 20:53:31 by doc^21o6 doc^21o6
Where did you hear about it?

There's a Canadian company called D-Wave who released a 16 qbits solid state quantum computer a few months ago, but it's not universal anyway. Apart from that there's Id Quantique in Switzerland and MagiQ in... US I think who are selling (almost) secure quantum cryptography systems.
Maybe they are selling fake quantum computers.
added on the 2007-11-09 21:02:01 by imbusy imbusy
I know of random number generators made by Id Quantique.. But real computers??? Not anytime soon, i guess?!
added on the 2007-11-09 21:16:02 by toxie toxie
in this http://www.heise.de/ct/ papermag.

i mean the same as you do. i've completely mixed it up.
and yeah, it's not universal, but i thought i'd post it anyway.
added on the 2007-11-09 21:22:07 by doc^21o6 doc^21o6
what's there to believe in, really? working quantum computers (if only very simple ones) have been around for some time now. the rest is a matter of time.

general purpose quantum processor: well, any quantum computation (with the models i know anyway) is necessarily invertible, which sure is a nice property for (cracking) cryptography or combinatorial optimization, but just as surely doesn't sound like a very useful property for general purpose computation (being able to throw information away sure saves on memory requirements :). i'm by no means an expert on the matter, but quantum computers seem pretty much dual to "regular" machine models (e.g. RAMs) in the sense that QCs can solve some problems a lot more efficiently than regular machines can, but would require extreme amounts of storage and computational capacity (boils down to the same thing in this particular case) to simulate a RAM. so it's not really a question of whether they'll replace current machines - if they ever get mainstream (and that still requires a bunch of really hard problems to be solved), i think it'll be in the form of add-on hardware for specialized problem solving (think dsp boards or GPUs here), not as a replacement for a normal cpu. (let's ignore for a moment that nobody really knows what a "normal" cpu will be in 10 years or so, and given the current state of affairs in quantum computing, 10 years to mainstream sounds extremely optimistic)
added on the 2007-11-09 21:22:13 by ryg ryg
BB Image

BB Image

heh the D-Wave Orion system looks and sounds very demo-ish :D
added on the 2007-11-10 12:08:14 by Zest Zest
except for some niche applications i guess conventional deterministic computing will remain dominant for a very long time. it took about 50 years for parallel computing to become common place, and it will take at least the same amount of time for quantum computing.
added on the 2007-11-10 13:17:24 by earx earx
i don't even know a demo that uses multicore :S
added on the 2007-11-10 13:30:06 by Zest Zest
well, as long as you're mainly pushing polys, there is near zero gain in multicore because gl/d3d interfaces are singlethreaded. and simulation/computation-heavy stuff is kinda tricky because you usually need a certain level of horsepower for it to run with proper speed at all. if you only get it fast enough using multiple cores, you just need them, and since there is that very vocal minority of morons who start huge flamefests whenever a 2007 demo doesn't run on their 2001 hardware, i can well understand that nobody wants to be the first to require it.
added on the 2007-11-10 13:45:01 by ryg ryg
That looks expensive. I want one.
added on the 2007-11-10 13:55:50 by doomdoom doomdoom
i don't believe in this newskool shite. (hey this feels elite)
BB Image
Awesome radiator !
added on the 2007-11-10 15:43:54 by willbe willbe
the D-wave stuff is *not* a quantum computer in the universally accepted meaning of the words "quantum computer".

And I'm with ryg on that part that if we manage to build working quantum computers, they will be probably pluggable into usual computers and designed to solve specific problems, at least for a while.
added on the 2007-11-10 17:16:49 by blala blala


Go to top