Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!


Forgot your password?

Bringing Quantum Chips To The Assembly Line 56

nutty_kong writes: "The National Science Foundation apparently is helping to develop a reliable way to manufacture quantum chips. So far, there has been experiments but no reliable way to reproduce or in other words manufacture the devices." Part of the problem with the promised magic of quantum computing is that component manufacture is often custom to each project, and difficult to repeat. Eventually, they'll hit RadioShack though ...
This discussion has been archived. No new comments can be posted.

Bringing Quantum Chips To The Assembly Line

Comments Filter:
  • by Anonymous Coward
    Killfile... first off, the singular form of Hertz is Hertz. It's a guy's name. Secondly, Hertz means cycles per second. This is a generic measurement that applies to just about anything electrical, not just computer processors. Therefore, I don't see it going away as a unit of measurement anytime soon -- and why should it?
  • Yeah, they're being produced but by the time you can see them you can't pick them up because they've entered a different quantum reality.
  • What I'm wondering is, how long will it be from the time we have quantum computers till one of them helps us design an infinite probability drive. Of course we know that every slashdot geek will get it and go "well the probability of me actually pouring hot grits down natalie portman's pants isn't 0, just very very small"....

  • Yeah, my first thought was, "all you really need to make one of these is a Brambleweeny sub-meson brain and a really hot cup of tea."
  • So does that mean if they know exactly how many they've produced, then they won't have the slightest clue where they are?

    Or is it that they won't know if they've produced the chip or not until they sell it?
  • Quick .. somebody drop this off with a large bribe to Tom's Hardware and get them to list it as better than AMD when it isn't!
  • Hertz is certainly not the be-all-end-all of chip speed. A 733MHz G4 can do lots better than some P4s out there, but lets not start a flame war over that.

    Usually you measure in MIPS(million instructions per second)
    Floating point Integers
    and GigaFlops and TeraFlops

  • Here a vendor is obligated to take back (refund) any device that fails. They give some kind of guarantee, most of the time it's about 1 year, but law actually regulates this to 2 years. Most ppl don't know this. After the first year you have to prove you haven't caused the malfunction yourself though.

  • It'll be great to bring quantum chips to the assembly line. Tying in with an older slashdot story, we can even ship them in rental cars [slashdot.org]. Sure, they'll know exactly how fast we're going, but they'll have no clue where we've taken their car!

  • ...is that, until you look at them, theres a 50% chance they wont be working.


  • "Radio Shack... You've got questions, we've got morons".
  • but have a very small performance increase over todays computers in searching ordered lists?

    That's because machines similar to Turing's can already bsearch() a sorted list in O(log n) comparisons. It's not like quicksort (a sorting algorithm with O(n log n) comparisons), where lousy implementations perform badly on already-sorted data, and some implementations fall back to Shell's gap insertion sort for ranges smaller than 10 or so items.

  • Normally prime factorization is an exponential (or "hard") problem.
    This is not correct. While prime factorization is "hard" in the sense that no one has found a polynomial time solution, it is not NP-hard. Thus, saying that prime factorization is an "exponential" problem, which implies that it is in NP, is incorrect. Reference: http://citeseer.nj.nec.com/aharonov98quantum.html
  • Pretty soon you will be able to go to Radio Shack to pick up your
    Schroedinger's CueCat

  • Ok they are talking about getting Quantum Chip to be able to substane room temp. , correct? Well, I have to say the recent comments on it release or lack there of is completely off the story? Any way i see they solving it with a liquid hydrogen solution if it doesn't shatter the processor or the motherboard just imagine just tapping your computer on your desk by accident .. opps i just SHATTERED my MAINBOARD!!!!!!SH*T lol
  • From the article, it seems that they need to develop a reliable method of laying down these devices atom by atom.

    If this can be done, then the implications for non-quantum devices are awesome. You could manufacture nano-machines, specialized conventional electronic devices, etc. If these techniques could be applied to the macroscopic world, you could create specialized materials with purities unmatched by conventional techniques.

    This technology bears close watching....
  • It's not surprising that these babies are hard to manufacture. I mean, making quality control observations must be a real bitch.


  • Yep, eventually quantum computing devices will reach radio shack...and fail three days after buying them, just like every other Radio Shack (TM) brand product I've ever bought...

    Me, I'm waiting for the day when I plug the quantum computer in my coffee cup in...of course that only works until I need a caffene fix: "Oh, Hell I drank my computer science project again!"

  • "At the time, the slashdot posters laughed, and moderated him down to (Score 0, flamebait)" "Three years later, under the iron boot of the King of England, noone was laughing."
  • I sounds a little like the "Infinite Improbability Drive." Maybe the Vogons can help.

  • no, no, no, its:
    "Radio Shack... You've got questions, we've got blank stares"
  • When the Kernel supports the chips... then I'll look into buying them.
  • um, it IS easily available from http://www.blackdon.org [blackdon.org] which you would know if you had actually bothered to read any of the articles on java for linux.
  • Why does this sound so much like the Planet Earth in Hitchhicker's Guide?...
  • I'm amazed that there's still a major retailer that you can walk into and actually buy individual transistors.
  • Quantum chips? Every flavor possible in every bag?

  • "Me, I'm waiting for the day when I plug the quantum computer in my coffee cup in..." Remember, they're trying to get this to work at room temperature. That's a nasty coffee. You want a nice piping HOT coffee, the kind that gets McDonalds sued, plug it into an AMD chip or any overclocked processor of the x86 line. I think I'll run out and patent that idea: CPU heat sink attached to a travel water heater. Great for International travel with your laptop. Jonathan
  • isn't it "I can almost smell the 30/60"?

  • *sigh* You guys just don't get this whole quantum thing...

    They might hit Radio Shack eventually.. and then again, they might not. Either way, you might be able to buy them, if not in Radio Shack, then somewhere else. Maybe.

    You're fucked if they go wrong, though, because you'll take them back to where you might have bought them and they'll say "Did you buy this product here, sir?" and you'll have to say "Well, I might have...", at which point, they'll say, well, it might not be our responsibility...

    Y'see, noone thinks about these practical day-to-day issues when they invent new technologies. They just do the press release and noone thinks about the logistics of manufacturing, shipping, stock control ("We might have run out of quantum chips.."), after-sales support, etc.


  • Wouldn't by quantum entanglemnet every chip be a beowulf cluster in itself?

  • Quantum Computing: Putting the N in NP!
  • Maybe Digital Convergence can jump on this and give away QuantumCues. All they ask is for a little processor time, otherwise its for you to play with.
  • And how is this different from the copy of Windows98 I bought last month?
  • Asked for comment, Sun's Scott McNealy responeded: 'WE put the "dot" in quantum-dot nanoswitches!'

  • You are correct. I didn't mean to imply that it was "hard" as in NP, but only that it is generally considered a "hard" problem by the mathematical community. The fastest known algorithm is exponential, and it looks to stay that way for at least quite some time.
  • Well if they were really smart, they'd only produce one chip, and then sell it to every branch of Radio Shack simultaneously.

  • I can see how they may be having problems in making the chips. In the same online Mag there is also this article:

    Tiny 'big bang' performs quantum computations [eet.com]

    Using a computer model that "explodes" a single particle into an infinite regress of quantum waves, University of Arkansas physics professor William Harter has demonstrated a new approach to quantum computation. "Our model reveals a fractal interference pattern emerging from quantum waves -- after what we are calling a tiny big bang -- that can perform useful calculations, such as calculating all the prime factors of any size integer," said Harter.

    Like I said, this could be difficult to manufacture into a chip.


    Check out the Vinny the Vampire [eplugz.com] comic strip

  • They might end up at RadioShack. Or anywhere else in the universe. Or nowhere at all. Or everywhere at once. Damn "probability wave" !
  • Eventually, they'll hit RadioShack though ...

    Yeah, right between the cm cubed terabyte of storage and the ream of colour electronic paper.

  • Aunt Millie will buy a quantum computer at Radio Shack and use it for sending e-mail.
  • Hate to burst the bubble here, but before the hype gets totally out of control, I think someone should mention that at the moment the best quantum computer has 2 bits, but will probably be scaled up soon. There are claims about one demonstration of 7 bits using NMR, but in my field the conventional wisdom is that this one doesn't do true quantum computing, and anyway it's not scalable. Unfortunately, the only working, 2-bit quantum computer, currently requires probably 1/2million$ of equipment to run, as it uses trapped ions in a very complex electrostatic trap inside a vacuum system. So far nobody even has a very promising proposal for doing quantum computing in a solid-state (i.e., really manufacturable) device, as far as I know, although when grant money beckons, people become intensely optimistic. On top of all this is the final point that there are only basically 2 known applications at which quantum computers would outperform classical computers: code breaking (remeber BB will have this technology long befor you do!) and the simulation of quantum mechanics problems (almost a redundant application, really.) The single-molecule, classical computer actually holds out vastly more promise for the industry.
  • Yes, and then if you feed it the improbibility ratio of a truely infinitely quantum processor you can create that out of thin air. You'll win the award for amazing cleverness and probably get lynched by all the other CS geeks who finaly realise that the one thing they can't stand is a smartass.

    Or is that tea?

    This has been another useless post from....
  • I can see it now.. Windows Quantum. Able to function correctly AND crash an infinite number of times simultainously so fast you won't even know it.... Hope my old DOS programs still run...
  • if (true) ... no, I mean if (false) ... no, I really mean if (true)..... Damn! It keeps changing every time I measure it....
  • One useful measure of a quantum computer is how many qubits you can maintain in a quantum mechanical superposition. The problem is called decoherence. The more qubits you try to link together the more likely they are to decohere from their pure quantum mechanical state, and behave more like a macroscopic object, in this case, say a regular old transistor bit.
  • by Gregoyle ( 122532 ) on Friday July 06, 2001 @08:26AM (#104699)
    You are ignoring Quantum Cryptography, where Quantum methods are used to encrypt data.

    That's what "unbreakable (by the laws of physics) encryption" is.

    Luckily, this looks like it will be possible much sooner than actual quantum factorization engines. It only requires one qubit, and has been successfully demonstrated. Hopefully someone will develop a method that does not need line of sight photon transfers. Although, when I think about it, that's integral to the encryption method.

    The whole thing with quantum cryptography is that it virtually eliminates the value of mathmatically based encryption algorithms. There is no present algorithm using superposition to generate even stronger encryption, and I'm not even sure if it would be mathematically possible to develop one.

    The only safe encryption methods will be the ones that use the properties of the quanta themselves. Hopefully someone will find a way to finagle a PKI out of that, or we go back into the Dark Ages of encryption.

  • So far one the most interesting use of quatum devices I have seen involved taking a wafer with an esentially random distribution of quatum dots and then using a neural net type arrangement to learn how to control it to get useful results out. I can't remember all the details but I think the setup was part of an optical light switching system.

    I thought it was an interesting way of getting use out of devices that we are still not able to produce with accuracy.

  • by kzinti ( 9651 ) on Friday July 06, 2001 @04:45AM (#104701) Homepage Journal
    Eventually, they'll hit RadioShack

    Oh, I can just see it now, hanging there on a swinging-door display, alongside all the other parts: RS part number 14-2847, in a little blue blister pack labeled 7404 TTL Hex Quantum Gate. On the back of the cardboard is printed a pinout and a Feynman diagram. I can almost smell the 60/40 now...

    Radio Shack: You've got questions, we've got blank stares.

  • by Dr_Cheeks ( 110261 ) on Friday July 06, 2001 @05:07AM (#104702) Homepage Journal
    A physics teacher back at my high school always used to blame any electrical experiments that he had problems with on stray capacitance - in 20 years time I'm looking forward to excusing segmentation faults as being an unavoidable result of the Heisenberg Uncertainty principle (and not just me counting from 1 instead of 0).
  • by TGK ( 262438 ) on Friday July 06, 2001 @05:53AM (#104703) Homepage Journal
    Does this mean we can start to look forward to the death of the Hert as a "measure" of chip speed? Since the ability of a chip to spin off quantum processes rather than vibrate will become it's dominant characteristic I'm thinking we might start to see chips measured by a different benchmark...

    Not that Hertz are really that terribly usefull or anything.... but they do give an idea, however vauge. Any thoughts on what new buzzword we'll be dealing with?

    This has been another useless post from....
  • by Masem ( 1171 ) on Friday July 06, 2001 @04:44AM (#104704)
    We no longer can say "Imagine a Beowulf cluster of these", because it will now be possible actually *have* a Beowulf cluster of these?

    Damn, those /. trollers are going to have to find a new tag line... :-)

  • by Tumbleweed ( 3706 ) on Friday July 06, 2001 @06:43AM (#104705)
    If Mac OS is ever ported to this technology, the conflict between it and Jobs' "Reality Distortion Field" may be enough to destroy the entire multiverse. Look at the bones, man!
  • by Monthenor ( 42511 ) <monthenor AT gogeek DOT org> on Friday July 06, 2001 @04:56AM (#104706) Homepage
    So I can know the speed of my computer or which process is running...but not both?
  • by Gregoyle ( 122532 ) on Friday July 06, 2001 @05:51AM (#104707)
    IANAPhysicist (at least not yet), but there are a lot of things that people, like the one who wrote this article, do not understand about quantum computing.

    First off; the "size" doesn't really matter. A lot of the article focuses on how the qubits will be XYZ times smaller than modern implementations of regular bits. Well, of course they will be!! The whole thing about quantum computing is that it uses the properties of a single atom; if the infrasturucture were much bigger than that it wouldn't make much sense.

    Almost as an aside, the article mentions the superpositioning of 1 and 0. This is HUGE . So what, you might say, what difference is it if a bit can hold a little more information? Think of this: Take 8 qubits. If these were normal bits they would be able to hold any one number from 00 to FF. When you have 8 qubits, they can hold ALL the numbers from 00 to FF. So you can run algorithms on all the numbers at once rather than just one at a time.

    Of course, it can only return one number at a time (meaning it might contain both numbers, but when you test to see which number it holds it will return one or the other). There are ways to get around this, though. In the mid 90s Peter Shor at IBM developed an algorithm [att.com] for prime factorization in polynomial time using qubits. Normally prime factorization is an exponential (or "hard") problem. RSA and almost all widely used public encryption algorithms rely on prime factorization for their security. This is important stuff.

    Some proposals for quantum computing use the "tunneling" method described in the article, but my hopes are with the NMR [stanford.edu] crowd. This seems the most promising using current technology.

    And as far as being able to buy this stuff at Radio Shack; I would be very surprised if that happened any time remotely soon. Think about it this way: unbreakable (by the laws of physics) encryption, and virtually instantaneous cracking of encryption, just for starters? Hmm, I can't think of any super-powerful world governments who would want to get their hands on that and keep it away from anyone else...

  • by Salsaman ( 141471 ) on Friday July 06, 2001 @07:16AM (#104708) Homepage
    You mean:

    if (true) printf("true\n");

    else if (false) printf("false\n");

    else printf("I am a quantum computer !\n");

The time spent on any item of the agenda [of a finance committee] will be in inverse proportion to the sum involved. -- C.N. Parkinson