Quantum Computing Demo Announcement

We have fixed the dates for the demo of our Orion quantum computing system. We are going to hold two events, one at the Computer History Museum in Mountain View, California on February 13th, and the second at the Telus World of Science in Vancouver, Canada on February 15th.

These events are open to the public, but registration is required as we’re expecting full houses at both events. Registration will open shortly and can be done via our website here.

The Orion system is a hardware accelerator designed to solve a particular NP-complete problem called the two dimensional Ising model in a magnetic field. It is built around a 16-qubit superconducting adiabatic quantum computer processor. The system is designed to be used in concert with a conventional front end for any application that requires the solution of an NP-complete problem.

Here is an optical picture of the processor we’ll be using for the demo. This particular circuit contains 16 qubits (the quasi-circular loops arranged in a 4×4 array). Each of the qubits is coupled to its nearest neighbors (N, S, E, W) and next-nearest neighbors (NW, NE, SW, SE) via a tunable flux transformer, giving a total of 42 of these couplers.

20070110_d-wave_orion_processor.JPG

Using an Orion system is extremely simple. I’m going to post more about this later, but qualitatively the way it works is that when your software application needs to solve an NP-complete problem, it passes the problem to the Orion system instead of whatever solver you’d be otherwise using. Nothing changes about how your app is architected.

At the demo, what we’re going to do is run two different applications, live, on an Orion system residing in Burnaby, BC. Orion is designed such that it can be used remotely, and this is the mode we’ll be using for the demos.

The first application is a pattern matching application applied to searching databases of molecules. This is an app that we developed internally at D-Wave. This is an example of how to apply Orion to problems arising from association (or conflict) graphs.

The second is a third-party planning/scheduling application for assigning people to seats subject to constraints. Anyone who has tried to plan seating arrangements for a wedding should be familiar with this one. This is an example of applying Orion to constraint satisfaction problems.

One very cool thing that we’re planning to do in Q2/2007 is to provide free access to one of these systems to people who want to either develop or port applications to it…so if you have an idea for an app that needs a fast NP-complete problem solver, start thinking about what you could do with some serious horsepower.

183 thoughts on “Quantum Computing Demo Announcement

  1. Hi Geordie,

    That’s very exciting news – I’m looking forward to hearing how it all goes.

    Can you tell us what the limits of the machine architecture as it currently stands are? From what I can gather, you can’t generate any 16-qubit entangled state that you please – is that correct?

    Can your processor run, for example, adiabatic versions of quantum search algorithms? What interesting quantum algorithms can your machine run besides these optimization-subject-to-constraints type algorithms?

    Thanks, and good luck!

  2. Joe: Thanks!

    Any problem that can be recast as a two-dimensional Ising model in a magnetic field problem (AKA quadratic integer programming) can in principle be solved using the approach we’ll be demo’ing.

    There are some quantum algorithms that can’t be run using the current architecture. The technical reason for this is that the devices that couple qubits i and j are of the \sigma_z^{i} \sigma_z^{j} type. There are some 16-qubit states that can’t be generated with the X + Z + ZZ Hamiltonian.

    There are some that can–for example the states |000…000> +- |111…111>.

    Our roadmap includes the addition of an XZ coupler to our architecture, which will make our systems universal. The reason for doing this is that we plan to build processors specifically for quantum simulation, which represents a big commercial opportunity.

  3. Hi Geordie,

    Thanks for the reply.

    When, according to your roadmap, do you plan to add this “XZ coupler” to the architecture? i.e. When do you expect to have a working 16-qubit universal quantum computer?

    You mentioned that by the end of 2008 you expect to have your system scaled to upwards of 1000 qubits. Will your prototypes with this many qubits be universal, or also of the restricted type you’re demoing next month?

    I forgot to ask last post: how much trouble do you expect scaling from 16-qubits to 1000-qubits will be? How long can you maintain coherence in your current system for? How do you expect decoherence to affect your 1000-qubit machine?

    Thanks,
    Joe

  4. OK, here’s a really crazy question, after reading Kaminsky and Lloyd’s Scalable Architecture paper. In talking about cooling the bits from infinity to zero, it says, “(physically, “∞” means “sufficiently high to make all possible states essentially equiprobable”)”.

    OK, so I start with random “thermalized” bits, set up the problem, cool the bits down to zero, read the answer. Now, how do I “warm up”/randomize the bits up again? Of course, under normal circumstances you’d just expose them to the rest of the universe. But this is a somewhat improbable universe in which the answer to the previous problem is known… and was computed by that computer.

    Is the nearby universe somehow correlated to the previously computed answer? Can it couple back to the bits, reducing their randomness? Or is there so much randomness (so many multiple-world choices, or whatever interpretation is fashionable) surrounding the nearest light-microsecond that you only have to wait one microsecond to couple the bits to something fully random?

    Or is this question based on an intuition so bogus that it’s not even wrong?

    Thanks,
    Chris

  5. Hi Chris: Good to hear from you. You may remember we met a while back at one of the Accelerating Change conferences.

    This is a good question. Whenever a “system of interest” interacts with “an environment” there is the possibility that information transferred from one to the other can “echo” back. For example, if I have two spins that are coupled, and I only look at one of them, its behavior looks periodic. If I couple three spins and only look at one, same thing but the period is longer. Etc. You can think of this period at the time it takes information about your starting conditions to echo back to where it started. I think it’s called the Poincare recurrence time in classical mechanics.

    In our case, the environment consists of many more degrees of freedom than the system of interest. In effect this means that whenever information is transferred from the quantum computing system to the environment, for all practical purposes it’s irreversible.

    Another thing is that we don’t anneal with temperature as you described (although this is a perfectly legit way to do it). We fix the temperature and anneal with a transverse field (equivalently by opening and closing windows for the qubits to tunnel). We never actively use a thermal environment to randomize the qubits, we just use the thermal environment as a bottomless pit into which we dump stuff we don’t need anymore. The analog of thermal randomization when you field anneal is to turn on a signal which cranks tunneling up to its maximum–I think this mode of operation is described in one of the Kaminsky & Lloyd papers.

    If you want a more technical explanation of the effects of thermal environments on systems like this take a look at the TAQC paper linked to in the sidebar under Publications.

    Hope this was helpful!

  6. Joe: The current plan is to first focus on successful deployment of the X+Z+ZZ type system. Many of the hard problems in operating a X+Z+XZ type machine can be resolved in our current approach. We already have a couple good XZ coupler designs, but we probably won’t start a processor line with these until the current line’s off the ground. Our roadmap has us introducing a quantum simulation processor line in 2009.

    About the scaling question: There’s no straightforward way to predict this. These systems are too complex and implementation-specific to know for sure. We will get to (and far past) the 1,000 qubit mark relatively shortly. Whether the systems will retain their current quantum behavior is unknown, but we haven’t seen anything that gives me concern about this. Part of this issue has to do with how the quantum computer is being used. The AQC approach is naturally shielded from errors in a way that the gate model isn’t, so whatever you’ve heard about error correction etc. take with a grain of salt, often these things are computational-model specific (although many people speak of them as if they were universal).

  7. Look, I am not aware of any theory that says that NP complete problems are amenable to any significant speedup on a quantum computer. (Factoring intergers, i.e. Shor’s algorithm, I remind you is somewhat special—it is not NP complete). In this case, you will not be able to compete with conventional computers.

  8. Another thing to keep in mind. The press conference method of announcing scientific results doesn’t have a very good track record

    In 1989, chemists Stanley Pons and Martin Fleischmann held a press conference to report they had successfully achieved cold fusion with a simple device.

    In 2002, a group called Clonaid held a press conference to announce they had successfully achieved human cloning.

    In both cases, the stories were widely reported in the press but were later debunked.

    How about some good old-fashioned peer review?

  9. And so what if you can find the ground state of a 16 spin Ising model. I’m willing to bet that in this particular physical device that quantum coherence has very little if not nothing at all to do with it.

  10. Donald:

    1. One of the most fundamental results of QC theory is that QCs can quadratically speed up unstructured search. I suggest you visit Eddy Farhi’s website at MIT and download and read some papers on AQC, or visit arxiv.org and search for adiabatic quantum computing. Most of the papers on AQCs are about solving NP-complete problems.

    2. We’re not announcing scientific results. We’re announcing a technical capability. When we do announce scientific results they will be via the peer review process.

    3. I would take that bet in a second, but unless you really are Donald Duck I would have difficulty collecting.

  11. 1. That is precisely my point. Quadratic speedup is not good enough to be competitive with current computing technology.

    2+3) Well, it’s not completely clear, but it sounds like you are claiming the technical capability to perform adiabatic quantum computation. If this is true you need to prove experimentally that what you have is AQC and not some sophisticated form of thermal annealing. This is what I would really like to see.

  12. Donald:

    I suppose if a quadratic speed up isn’t good enough, then a constant pre-factor speed-up must be even less useful…damn thanks for pointing that out…now I can go back to using my trusty ole abacus. You should probably email Intel and AMD and let them know. Damn “computers” and their useless pre-factor speed-ups.

    I understand that presentation of scientific results in Science or Nature is appealing to the expert community, and we do have plans to do this. But our primary objective isn’t publishing science papers, it’s building quantum computers.

  13. So will this help me find songs on my iPod more quickly? :)

    It’s so cool that you’re demoing this . . . since I can’t make it I’ll keep an eye out for on youtube. Any chance you’ll be putting out a programmer’s guide? Programming Orion for Dummies? :)

  14. Geordie:

    True, quadratic speedup for general purpose computing would be nice—if the cost is not too outrageous. But that’s not what we are talking about here. AQC may give quadratic speedup for a few select algorithms, e.g. Grover’s search algorithm. There are also problems known to be exponentially hard using AQC. I think its very much still an open question as to how useful AQC is w.r.t. computing in general. Yet I also think that studying this will pehaps tell us something very fundamental about the nature of computing and possibly physical reality. However, I’m not convinced that there is now, or ever will be, a market for AQC.

    Back to your device. I read somewhere else that your technology works at -269C, i.e. 4K, so I take that to mean a liquid Helium temperatures. Now from what I hear, individual s.c. flux qubits, including yours, have a energy gap E0 of about 10GHz or 0.5K. My guess is that a modest collection of coupled flux qubits as in your ‘processor’ has a minimum energy gap ~2 orders of magnitude smaller than E0. So the temperature is something like 3 orders of magnitude greater that the minimum energy gap. How is AQC possible here? How can you even initialize the system?

  15. The only way that I would believe in DWave is:

    (1)DWave publishes a precise description of how to write software for its device, including all limitations
    (2)DWave pays an independent company to write a simulated annealing program that solves the same problems, with the same accuracy, using state of the art distributed computing with 100 conventional computers
    (3)DWave shows that their machine can get a solution in 1/100 the time than (2).

  16. Dwave is demoing the solving of the wedding seating problem and the pattern matching problem. Their 16 qubit machine is not yet faster than regular computers. 16 qubits is 64000 states.

    When they get up to 64 qubits or more then we should start seeing better speed. But since the plan is to scale to 1000 qubits or more in 2008 this will happen very soon.

    Note: other quantum computer demos with other techniques were adding two numbers.

    Dwave does not make money until they are solving the problems a lot faster.

  17. Hi rrtucci:

    We do plan to provide (1) very shortly, in order to show people how to use the system we’re freely providing for apps developers.

    As for (2) and (3), the current system won’t be able to provide that level of performance. As Brian suggests, this very much is a proof of concept. It works very well, but the qubit array is too small to allow the system to compete with conventional approaches.

    However as the arrays get larger, the programming interface will remain identical–the only thing a user will notice is a big speed-up from generation to generation of improving hardware.

    If we do our job, the system will catch conventional approaches shortly, and we’ll be able to supply your (2) and (3) (and a lot more!)

  18. Donald:

    There are only two reasons why QCs will ever be built: quantum simulation and solving NP-complete problems. Both of these represent enormous markets. We’ve checked.

    Re. your questions about temperature: these are excellent questions. As a generalization of your question, think about ANY AQC operating on a “hard” (ie exponentially small gap) problem. Is there any physical system whose temperature is smaller than the gap at an anti-crossing of a hard problem? Of course not.

    All AQCs have the feature you’re describing, not just our approach. At an anticrossing, the temperature is ALWAYS going to be orders of magnitude larger than the gap. That’s why inclusion of a thermal environment is REQUIRED in order to analyze how to operate an “AQC” (although note that at the anticrossings it’s not really adiabatic anymore).

    In order to see what happens when T>>\Delta take a look at the TAQC (Thermally assisted adiabatic quantum computation) paper in the sidebar.

    Qualitatively, the effect of the large temperature is to thermalize the two energy levels involved in the anticrossing, reducing the probability of success by 1/2, which is of course completely acceptable.

  19. The unfortunate reality is that this is really just classical SFQ being used for what is effectively analog computation (i.e. system simulation). The fact that only Z coupling is achieveable attests to this. Further, given that nowhere in any of your discussions does DWave ever mention quantum coherence, T2, phase evolution, or superpositions, one is forced to believe, as I said, that this is effectively a classical machine.

    Frankly, you really shouldn’t call your SQUIDs qubits, as they are no more qubits than are the SQUIDs in SFQ pulse generators. They are two level systems (clockwise and counterclockwise propagating persistent currents), but the quantum nature of said system is never exploited!

    Indeed, given that all experimental results to date have shown coherence times of order ~10-100ns for Nb trilayer devices, I’d be shocked to learn that Dwave had somehow overcome this technological hurdle ahead of the entire research community. If I’m incorrect, please publish something demonstrating quantum coherence using your “qubits” and prove me wrong. I’d be thrilled with such a response.

  20. Scrooge:

    ::sigh:: OK I understand that for some reason you’re desperate to find some reason why what we’re doing can’t possibly be correct, which is fine.

    I’m familiar with this approach. It goes something like this: I can’t figure out how to do it, therefore you can’t figure out how to do it.

    Do you want me to point out the basic flaw in this reasoning or can you figure it out all by yourself.

    As to your specific comments:

    There is NO SFQ in this design. Zero. The qubits are compound junction RF squids. The tunneling matrix elements for each qubit can be controlled by varying the flux applied through the CJJs for each qubit. This approach is well-known and is centrally featured in the superconducting AQC papers I’ve linked to here.

    As I mentioned earlier the Hamiltonian is of the X+Z+ZZ type. Notice the X?

    As to your comment that I haven’t talked about T2 etc. As you yourself pointed out scientific results belong in peer-reviewed scientific articles, not in a blog whose objective is to reach a broad audience with a message that isn’t completely incomprehensible because it’s buried under jargon.

    As I said before, our objective is to build quantum computers, not to publish science papers. If the latter supports the former, we’ll publish. If it doesn’t then it’s just a distraction for us.

  21. Geordie,
    I did not claim that you are using SFQ, I claim that the behavior of your system is akin to classical SFQ. My apologies if the word choice was confusing.

    My criticism of your approach has nothing to do with me figuring anything out, or an apparent claim that I have been unsuccessful in doing so. I don’t work in superconducting qubits. However, I know the field, and the MANY MANY players as well as the challeges they face. You are claiming to have surpassed them all by more than an order of magnitude in the number of qubits you can control and manipulate. Such a claim warrants a publication, or a detailed press release, or something to suggest that you have actually just ushered in the computing revolution which you are claiming. You may not be in the business of publishing science papers, but you are in applied science. The validity of technical claims in ANY applied science discipline is upheld by scientific scrutiny, generally facilitated by publishing scientific results. Would you prefer a webinar? Fine, but demonstrate the behavior you are claiming transparently for all to see.

    Further, you shouldn’t fall back on the fact that this is a blog. I have read DWave’s papers on the arxiv and find the same lack of anything quantum coherent in your published results (e.g. cond-mat/0509557, cond-mat/0501085). Dwave and collaborators certainly know how to make quasi-classical superconducting electronics and SQUIDs, but where are the superposition states? the Rabi or Larmor oscillations? anything suggesting that you are operating and controlling a coherent quantum system?

    I understand the premise of AQC, but again ask this: Can Dwave demonstrate that their simulator/processor can take an input superposition state and output the appropriate answers in superposition? If so, please provide the data and I will be most impressed and GLADLY give you the credit you are due. I

    n stark contrast to your claim, I am not desperate to find some reason why what you’re doing is incorrect. Nothing could be further from the truth, but I do expect reasonable experimental evidence to support your very significant claims.

  22. Scrooge:

    Fair enough! While we obviously can’t release everything we’ve learned from the hardware, what we’re planning to submit for publication should clarify (at least) the issue of the role of QM in the operation of the system.

  23. I’m looking forward to those publications, but have a follow-up question. Your statement that said publications will “clarify the role of QM [quantum mechanics] in the operation of the system,” gives me pause. We understand the role of quantum mechanics in quantum computing; does the DWave system exploit QM in the same way? Or are the effects what one might term semi-classical? For example, QM plays a significant role in the operation of the laser, the FET, and classical SFQ logic, but none of these are coherent quantum devices. By this statement I mean they do not preserve and exploit quantum mechanical phase information. Accordingly, they cannot provide the parallelism which leads to exponential speedup in a quantum computer. How would one describe DWave’s system?

  24. @Scrooge

    Quick question. How does one read out a state in a superposition? Is it performing repeated computation and seeing a 50-50 chance of two results? Or is it computing in one basis and reading out results in an orthogonal basis (i.e. operate in|x> read out in |z>)….just curious.

    @Geordie

    Im going to be at the Vancouver presentation and I was wondering if you would be going over technical aspects of the initialization and readout of your qubits, or if the presentation will focus more on the running of the computations. I am still scratching my head as how after the computation is run, you read out and interpret your solution…I think my problem is I am still thinking about a binary input/output, when in reality I dont think, with the analog nature of your system, that is how your system runs. Thanks again!

  25. Scrooge:

    I am not so sure you’re correct when you say that the role of QM in QC is understood. There are of course lots of things that are known, but there is still alot of unexplored territory. The example you brought up about temperature & the role it plays in AQC is a great example. From the theory perspective, adding environments qualitatively changes the behavior of the system. I don’t believe that even this simple point is widely understood. There are lots of things like this where computation and physics are related in non-trivial ways, and where cross-overs between classical and quantum behavior may affect computational scaling in a way that isn’t just either/or.

    Also just to be clear I don’t believe that the system we’re building is going to exponentially speed up anything. The objective is the quadratic speed up for unstructured search.

  26. Chris (and also Scrooge):

    The way we operate our AQCs is like this (X_i and Z_i are the pauli X and Z matrices for qubit i):

    1. Turn up the tunneling term in the Hamiltonian to its maximum value (H=\sum_i \Delta_i X_i)

    2. Slowly turn the qubit biases and coupler strengths up to their target values (these define the particular problem instance); after this process the Hamiltonian is H=\sum_i (\Delta_i X_i + h_i Z_i) +\sum_{ij} J_ij Z_i Z_j

    3. Slowly turn the tunneling terms off; after this the Hamiltonian is H=\sum_i h_i Z_i +\sum_{ij} J_ij Z_i Z_j

    4. Read out the (binary digital) values of the qubits

    OK so the point of this is that the qubits are only read out when they are in classical bit states by design. The readout devices are sensitive magnetometers called DC-squids which sense the direction of the magnetic field threading the qubit and hence it’s bit state. The computational model is explicitly set up so that superposition states are used only during the “annealing” stage; the readouts never fire during this step.

    Answers are encoded in bit strings. Each bit string corresponds to a particular solution. If the computation succeeds, the bit string returned ({s_i}) will minimize the energy E=\sum_i h_i s_i +\sum_{ij} J_ij s_i s_j.

    Hope this helps!

    Also re. the demo. There will be almost zero technical stuff in the demo. The foxus is on describing how one would use the system as an application developer–what it does and how you interact with it. All of the science-type stuff, including details of operation, won’t be part of the demo.

  27. Hi everybody: As a favor to our non-technical audience, if you have any technical questions about the system, please email me directly at rose@dwavesys.com and I’ll try to help.

    Also Donald and Scrooge: Sorry about cutting your posts, please email me directly & we can continue the discussion.

    I love the feedback, keep it coming!

  28. As a quantum computing layperson, I agree that most of the jargon is out of my reach. But it would be nice to know what the key points of the technical discussion are. . . where does your quantum computer derive its power: entanglement, tunneling, etc.?

    It doesn’t seem that the world (at least in Google’s view) has yet taken notice of your announcement. What I mean by this is that skeptical experts might not need to worry about any damaging press to the QC field should D-Wave’s claims be false. I know this likely won’t be the killer app, but if a press release had been issued indicating plans to demo an RSA code breaking quantum computer, then I could see cause for worry.

    I’m an optimist, and can’t wait to see more. Go D-Wave!

  29. “The objective is the quadratic speed up for unstructured search.”

    Come on Geordie, shoot higher! I mean adiabatic quantum computation IS “equivalent” to the standard circuit model. If you’re going to be “crazy D-wave” why not go all the way!!! :) :)

    (Unfortunately with a ZZ+Z+X model, I only know how to do the single qubit equivalence…hmmm….now that’s a good problem! Oh wait…oh, I see how to do it.)

    Oh, and sorry about the technical post.

  30. Dave:

    There’s this guy in the WWE called “Super Crazy”. His shtick goes something like this:

    “People ask me why I’m called Super Crazy”. Pause. (also insert mexican accent). “Eets beecoze I’m SSSUPER.” Long pause. “And I’m CRRAZYY!”

    For some reason your post triggered that. Sorry. Ahem.

    Anyway the reason why we don’t think we can run quantum simulation (the only useful algorithm with an exponential scaling change) in general is that X+Z+ZZ isn’t universal for BQP. There are states you can’t get to with a Hamiltonian like this. Also the only problem with a proof that our computational model (with noisy qubits) provides the same scaling as a fully coherent gate model version is unstructured search.

    It’s almost certain that as we get better at fab, reducing noise in electronics, theory, new and better device designs, etc. that this approach will lead to full BQP solvers but we’re a long way away from this. One step at a time!

  31. @Tom

    Geordie did a pretty good intro to quantum computing on this blog. I’d start with “The basic idea behind quantum computers” and read the posts he did at that time. A link to it is in the widget bar. Hope that helps.

  32. Geordie, thanks for your answer. I knew you were using magnetism instead of temperature (I did read the paper!) but I thought the two were equivalent in theory and I could express myself more clearly in terms of temperature.

    I’d also like to encourage slightly more technical talk on the blog. As a programmer without much familiarity with quantum systems, simply knowing how to program the system is good enough for me–but that requires that I know a bit of the causality involved.

    To program the system, I’d have to be able to think through the system’s state transitions and see why the biases and couplers have the effect they do–why a certain pattern leads to a certain answer–and how slowly I have to change the values. Once I know that for toy problems, I can generalize.

    It sounds like this is what you’ll be explaining at your demo?

    I started to say that the answer to my question should satisfy the more physics-focused people as well. But by analogy, telling me how to program a logic circuit would tell me nothing about whether the transistors were FET or NPN.

    And in fact, if I understand correctly, knowing how to program your system–and seeing it get the right answer for small problems–won’t tell us whether you’re doing classical annealing or actually maintaining superpositions. Suppose your sixteen-bit system, when “thermalized,” tries 1E12 combinations per second, classically (based on a guess about lightspeed across the size of your system). That’s enough to find the right answer randomly in a millisecond or so. A thousand-bit system might try 1E11 combinations per second (due to its greater size) (and I’m just guessing that it scales with linear size–it might be worse). That’s not nearly enough to cycle through the combinations classically.

    So I suppose I can’t blame Donald and Scrooge for being skeptical (though I found their tone impolite). I mean, I assume you’ve thought this through, and I assume you’re not being fraudulent. But someone who was inclined to be skeptical might continue to be skeptical even after seeing the 16-bit system solve problems… if I understand correctly how a merely-classical system might mimic a quantum system at small scale.

    Thanks,
    Chris

  33. Chris:

    “To program the system, I’d have to be able to think through the system’s state transitions and see why the biases and couplers have the effect they do–why a certain pattern leads to a certain answer–and how slowly I have to change the values. Once I know that for toy problems, I can generalize.”

    To just use the system you don’t need to know this. What you’re describing is akin to programming in machine language for a conventional system–thinking about software in terms of the sequences of voltages applied to certain devices etc. While of course we have to know how to do this to operate the machines, a user doesn’t need to know this (and generally won’t want to).

    For example, one way you can use the current machine is to provide a graph. The system goes through a series of steps that outputs the Maximum Clique of the graph. As an application developer, all you need to know is what format to send the graph in and what format to expect the answer back in. None of the “under the hood” stuff is required reading.

    Some users will want full access to the machine language, and we plan to offer full use of the machines through this level of abstraction. However nearly all commercial users expect plug n play, so the system is architected to be as close to a “black box solver” for application development as it can be.

    “And in fact, if I understand correctly, knowing how to program your system–and seeing it get the right answer for small problems–won’t tell us whether you’re doing classical annealing or actually maintaining superpositions.”

    Yes of course this is true. My view of this is that ultimately the only thing that matters to a user is the wall-clock time it takes to get a satisfactory answer. If the system really is a QC (which all of our internal testing clearly shows is the case) as it matures we’ll be able to calculate many orders of magnitude faster than the best known conventional approaches.

    If for some unforeseen reason as we scale the system up to kqubits and Mqubits the QM we’re using goes away, we’ll only have an ultra-fast nearly zero power massively parallel NP-complete problem solver, and we’ll only be able to do a few orders of magnitude faster than everyone else :)

  34. hi Dave!

    on the ZZ+Z+X model: “single qubit equivalence”? I don’t quite understand it, since the transverse Ising model shows tunneling and hencefore quantum behaviour…

    hi Geordie!

    on the ZZ+Z+X model: do I understand correctly, what one can do unstructured (Grover-like) search with ZZ+Z+X providing quadratic speed-up?

    Thanks,
    Ron

  35. Hey Ron, Geordie,

    Again apologies up front for my technobabble. But I agree Georgie, small steps! Of course small steps have a habit of becoming leaps before you know it!

    To clarify there are two approaches one can take. One is to try to use the X and Z and ZZ as “gates”. This is indeed universal, but you need to do some encoding…(not also that if you only have ZZ on nearest neighbors on a line this is not universal. Details about this can be found in my thesis, in work by DiVincenzo and Terhal, and in work from the Whaley group after I left and others I’m surely forgetting.)

    But what I was thinking about was something different. Recently Aharanov et al plus a host of others have shown how to build a quantum computer using adiabatic quantum algorithms. They do this by showing how you can construct an adiabatic quantum algorithm whose final (ground) state is the HISTORY of a quantum computation. Then you measure the HISTORY. This works because you can pad a quantum circuit by identity gates and then you end up with a high probability getting the outcome of the circuit when you measure this history. The question I was thinking about whas how to use ZZ+Z+X as adiabatic interactions and build such an architecture. Its fairly simple to build a single qubit version of this, well you end up with only real gates, but that is fine. Going to more qubits is difficult…I think I see how to do it but need to do some number crunching.

  36. Chris,

    Thanks for the info.

    I think it’s great that this will be opened up for trial use by early-adopters. It would be neat to see some real hardcore combinatorial optimization applications that traditionally use heuristic techniques to find solutions. Problems like circuit design or antenna design, or any (mechanical) structural optimization problem.

  37. Rose said
    “Yes of course this is true. My view of this is that ultimately the only thing that matters to a user is the wall-clock time it takes to get a
    satisfactory answer. If the system really is a QC (which all of our internal testing clearly shows is the case) as it matures we’ll be able to calculate many orders of magnitude faster than the best known conventional approaches.
    If for some unforeseen reason as we scale the system up to kqubits and Mqubits the QM we’re using goes away, we’ll only have an ultra-fast nearly zero power massively parallel NP-complete problem solver, and we’ll only be able to do a few orders of magnitude faster than everyone else :)”

    These two paragraphs are mere speculation and Walter-Mitty-like day dreaming. Dwave Investors should demand to see something like (2) and (3) that I have outlined in my previous post.

    Sound business practice (common sense really) dictates that one should compare a device one is investing millions in, with its nearest competitor in the marketplace. In this case I would say that competitor is distributed computing. Dwave doesn’t seem to be willing to do this comparison in a quantitative way, except in the far indeterminate future. This bodes badly for DWave.

    The pictures of the electronics are cool, but just because the machine looks cool does not mean it will make a profit. It’s not hard to put beatiful covers on a bad book

  38. rrtucci:

    “These two paragraphs are mere speculation and Walter-Mitty-like day dreaming. Dwave Investors should demand to see something like (2) and (3) that I have outlined in my previous post.”

    UUggg where do I start.

    OK how about here: Of course we do systems profiling. I can’t comprehend what would lead you to think we didn’t. Isn’t the whole point of what we’re doing increased performance? Don’t you’d think we care just a bit about the only reason for doing this? We know EXACTLY how the current system compares to other approaches.

    We also know, and I’ve told you this before but you don’t seem to be listening, that the current system is NOT COMPETITIVE with state-of-the-art conventional approaches . It’s a systems proof of concept, no more and no less. As the hardware matures, the system architecture (and most importantly the user interface) can remain fixed–the only thing that changes is a chip swap-out.

    Another thing: calling a well-planned and well thought out roadmap speculation is misleading. In a sense any forecasting is speculation. Does that mean that plans are worthless? Of course not. We haven’t kept our investors happy (which seems to be a particular point of interest for you) for eight years by missing milestones.

    “Sound business practice (common sense really) dictates that one should compare a device one is investing millions in, with its nearest competitor in the marketplace. In this case I would say that competitor is distributed computing. Dwave doesn’t seem to be willing to do this comparison in a quantitative way, except in the far indeterminate future.”

    Your first statement is obviously correct. We spend a lot of time understanding other approaches to these problems, and not only for the reason you’ve raised. Your last statement is false.

    Remember it may not be a top priority for us to phone you up and tell you everything we’re doing.

  39. Dave:

    My understanding is that X+Z+ZZ type Hamiltonians (even in the gate model) can’t run the general phase estimation procedure regardless of encoding. Am I wrong?

    Answer: I have obviously fallen on my head a few too many times.

    Sorry momentary wires crossing moment.

    I think what I meant to say was that the equivalence between AQC and universal gate model stuff requires terms in addition to X+Z+ZZ in order for the method you were describing (Aharonov etc.) to work. Although I have been in an airport 7 out of the last 22 hours so I’ll revisit tomorrow :)

  40. Geordie,
    Are both events going to be filmed and posted on the DWave site? Any chance of a webcast? Im going to the Vancouver demo, but I am curious to here what questions and comments are raised from the brains down in Google-land. Thanks again!

  41. Pingback: The Quantum Pontiff » Orion Into The Future

  42. SInce you are posting the exact location of the computer, does this mean that you do no know how fast it will be ?
    And is the other computer whose speed is know the one that will be either in Los Angeles or Santa Barbara (but is still undecided) ?

    nb: I guess that now that quantum computer seem to really exist I’ll have to really read about them and try to understand what is quantic about them.

  43. Hi there. That’s a pretty amazing piece of technology that you and your collaborators are assembling.

    The technical skill in how to filter, heatsink and wire many lines into a dilution fridge is going to have a value in the marketplace irrespective of the final application.

    I wonder what heat load you are achieving at the mixing chamber per control line, and whether you have projections for how this will change with time?

  44. Hello there. I am to start university a year after this term ends and I am looking to be doing Computer Science. But at the same time, I am highly interested in quantum physics and I thought that quantum computing would be a very suitable subjects for me. I’d like to start reading about it but I don’t really know where to start and I wanted to know if there are any courses at university about it yet and how it is able to be studies alongside computer science. Thanks for your time.

  45. Pingback: blogspot.sk » Blog Archive » D-Wave Systems 16-qubit quantový počítač

  46. Pingback: scalability.org » Blog Archive » P=NP and other trivia

  47. I understand that QC is applicable to NP-type problems such as decryption, pattern matching and scheduling, but how does it fare in P-type applications such as FFT? ie, would a QC operate faster than Von-Neuman Architecture in this application? And more importantly, can you program your device for a faster FFT?

  48. 3. I would take that bet in a second, but unless you really are Donald Duck I would have difficulty collecting.

    It’s simple – search for DD in the quantum realm :)

  49. … even though spacecraft are not yet ready for it, how about plotting return flights to & from all the solar systems within … say … 50 light years …

    … vast amounts of contraints to consider along with a whole lot of numbers to crunch …

  50. Pingback: Quantum computer to debut Next Week

  51. Pingback: Quantum Computing Demo Announced « PoliTech

  52. I understand that for the time being the hardware incarnation of your design is limited in terms of its capabilities, but I’m wondering whether you’ve considered the applications of QC to AI. For instance, a QBit (or an array of them) could be used to represent the weights of a Neural Network, and the increased search capabilities of QC could speed up the process of finding a set of weights which encode a particular function. Of course this would require a large number of QBits, but I see that’s part of your plan ;) There could be a wide array of applications of such a technology, and I’d be interested to hear your thoughts on the subject. I look forward to attending your demonstration. Lucky thing I happen to live in Vancouver.

    Jared

  53. Pingback: Hi-Tech News and Gadget World » Quantum computer to debut next week

  54. Pingback: TechChee.com, online shopping for hot tech stuff » Blog Archive » Quantum computer, the magician in the computer industy

  55. Pingback: informatica.ingenierosweb.es » Orión: avances en la computación cuántica

  56. Geordie,

    re – “If for some unforeseen reason as we scale the system up to kqubits and Mqubits the QM we’re using goes away, we’ll only have an ultra-fast nearly zero power massively parallel NP-complete problem solver, and we’ll only be able to do a few orders of magnitude faster than everyone else”

    Is the success of the scalability (to more useful array sizes) dependent on the coherence length? and if so is the CL possibly intrinsic to the wiring characteristics? And more specifically grain nano/microstructure in the niobium affecting or limiting the coherence? Is this what you might be alluding to with the phrase “unforeseen reason”?

    If the grain structure is at root cause to the limits of CL in the interconnect, there are well known PVD (physical vapor deposition) methods that can forcibly vary thin film grain structure quite substantially – although not likely common in JJ niobium Thin Film deposotion.

    Not merely garden variety sputtering conditions btw, as there are more subtle techniques involved, often using different apparatus than common sputtering.

    If the wiring itself limits the CL, have you fabricated test structures of wiring (lengths widths and thicknesses ) that can help you predict array scalability / CL improvements to some degree of confidence, and possibly measure the CL to determine effects of process variables on grain structure, and on ultimate improvements to CL?

    I sense that from reading a bit about your efforts you might not be assessing the CL issues ?possibly directly related to the materials effects and how possibly to manage/ optimize these more directly for CL improvements (and ultimate scalability).

    This might be something to consider if presently you treat the foundry as a black box, as success in improving CL may have a strong impact on your product scalability and ultimate commercial success. I doubt that mere sputtering power or pressure will do what you ultimately wish, although there might be ever so marginal effects.

    Additionally if this remains true of planar Niobium wiring test structures, I’d hazard a guess that there are similar effects in the grain structure effects traversing across thin film topography, which might also bear some reflecting upon as to possible improvements, aside from mere deposition conditions and methods.

  57. So, I’m just at simple ordinary everyday person, and I’m wondring what effect this quantum technology will have for people like me? What can I expect from the future? Central computing services? Faster personal computers? Super realistic realtime gameworlds? Sentient Robots? or are the quantum chips only for the sience community? Sorry for the stupid question. Oh, and one last thing, how much faster then ordinary computers is this thing, if comparable at all?? thanks for your time :)

  58. Pingback: Nice Ventures Blog

  59. Pingback: meneame.net

  60. Pingback: Quantum computing? « KuiperCliff

  61. Pingback: www.romanding.ch

  62. Could you explain the theoretical difference between your system and the operation of a classic CNNA circuit? (analog cellular neural network) CNNA-s can be used to solve partial differential equations or for pattern matching. (the biggest system that publicly exists today is around 128×128 nodes and used for real time pattern matching) My second question would be if it’s possible to run the same calculations your system is performing with a CNNA chip and quantize the result with the binary logic stage of the chip and get the same results?

  63. Pingback: Tech Suki » Blog Archive » Quantum Computing Demo Announcement

  64. Pingback: Open Spaces » First quantum computing demo scheduled next week

  65. Pingback: Quantum Computing Test « V.276.se

  66. Pingback: שווה קריאה

  67. I for one wouldn’t give a rodent’s buttox if a “non-quantum computer” just happens to solve problems faster than any “quantum computer” (or “quantum method A computer” versus “quantum method B computer” for that matter) that anyone else could cook up at this point. In practical application any Joe off the street would know that performance matters more than a shinny sticker that says “QC APPROVED”

    Go to a university for machines that run exactly how papers defined them to run, and go to companies for the actually useful ones

  68. @John
    I believe the system is being operated remotely because it takes up a huge amount of space and needs proper EM shielding. This thing ain’t the size of a desktop. However, you will see posted on this site many papers outlining DWaves technology.

  69. Talk about jumping the gun. If this pans out the accelerating returns and singularitarians are gonna be having epileptic fits of epic proportions. I mean.. imagine how computational power like this will factor into protein folding and drug development! I really am going to live forever!

  70. Pingback: Quantum computer to debut next week at BlogThe.Net

  71. Pingback: First Quantum Computer? « Words

  72. Pingback: 16 Qubit Quantum Computer to Debut on Feb 13th; 20 Years Ahead of Schedule -Simple Thoughts - Java and Web Blog

  73. Pingback: The Fine Art of Eccentricity · Tom Cruise Wants to be Funny. Ha.

  74. Another practical aspect is that high count multicore cpus for mainstream markets are going to become considerably more prevalent and I don’t mean dual core or quad core.

    Both the recent GP-GPU thrust and the integration of sigificant numbers of vector FPU execution units into multicore single chips is looming. With present discrete GP-GPUs single precision FP performance is up ~10x over conventional CPUs.

    With Intel’s multicore Larabee in the works, (see
    http://www.theinquirer.net/print.aspx?article=37548&print=1 )

    Intel will be generating 16->64 short pipelines cores + VECTOR FPU units in single chip configuration (min 16 core).

    There might even be multiple Vector ALUs per core…. Increase the net Vector FPU per CPU socket by say 16x cores, say executing 16 long vectors, and there is a slightly different horizon to aim for to compete for computational power of the future cluster CPUS… say 1000 Gigaflops at 4Ghz…on 1 chip…Do that in a large cluster network multiprocessor and I dare contemplate the performance.

    Larabee won’t compete in NP complete problems unless CL becomes a problem for scaling the SC QC… but it sure is interesting how fast computational power will increase even conventionally.

  75. Pingback: Quantum Computer To Launch Next Week « Alex Torex Blog

  76. Pingback: 16-qubit quantum computer - Cool Sci-Fi Forums

  77. Pingback: » dxdt.ru:

  78. Demo Suggestion
    There is a general problem that could benefit from a quantum approach…linking stories to numbers as might happen when someone wants to know the implication of a world event to their business.

    How does the story affect my revenues, costs or profits?
    What might be the effect on my customers or suppliers?

    Look forward to your February events and I hope to be in Vancouver in April or May.

    Cheers,

  79. DD saids
    “Another thing to keep in mind. The press conference method of announcing scientific results doesn’t have a very good track record

    In 1989, chemists Stanley Pons and Martin Fleischmann held a press conference to report they had successfully achieved cold fusion with a simple device.

    oh and on cold fusion
    Despite the ridicule, there has remained a dedicated and growing core of scientists who have not only replicated room-temperature fusion, but also have improved on its performance and have broadened the number of methods for achieving it.

    Also
    Researchers at Rensselaer Polytechnic Institute have developed a tabletop accelerator that produces nuclear fusion at room temperature, providing confirmation of an earlier experiment conducted at the University of California, Los Angeles (UCLA), while offering substantial improvements over the original design.
    at

    http://www.scienceblog.com/cms/ny_team_confirms_ucla_tabletop_fusion_10017.html

    so it looks like something is going on

  80. re: Scrooge [We understand the role of quantum mechanics in quantum computing; does the DWave system exploit QM in the same way? Or are the effects what one might term semi-classical?]

    From a pragmatic perspective — I don’t care about the true underlying physical model for how it works as long as it produces accurate results. Whether the reality is AQC, some other QM, or some other as yet to be determined property of the universe doesn’t matter in the short term … reproducability of results does matter. In the long term, it would be great if we learned the true nature of its functioning and paved the way for even more profound discoveries about the fabric around us.

  81. joep said:
    “oh and on cold fusion

    Researchers at Rensselaer Polytechnic Institute have developed a tabletop accelerator that produces nuclear fusion at room temperature”

    There’s nothing cold about pyroelectric fusion; indeed, heat is the input energy to the crystal in order to create the accelerating potential.

  82. Pingback: Orion quantum computing system « asking answers

  83. Pingback: Kryptoblog » Blog Archive » Demonstration av kvantdator

  84. Geordie,

    Many thanks for all your patience here.

    In Detroit they understand “proof of Concept”:

    http://www.edmunds.com/news/autoshow/articles/100916/article.html?mktcat=autoshow&kw=concept+car&mktid=ov204246

    So I understand what you are doing and the ones so tied to their own preconceptions of what a quantum computer should be are stuck in a quagmire of their own making.

    Proof of Concept says – we can make this happen and here is one way it could look. Then the manufacturers run with the design and find ways to solve problems not at first obvious. When people were feeding hand punched data cards in one at a time to Factory sized computers did people ever dream you could interact as we do now with computers?

    This is the same idea. You start with something and where it ends up may be in a totally different direction. To me a business introduces a new product and researchers write papers. Your company has made a product and you say this is what we have done and we can do even more later. Makes sense to a business but not to an academic.

    VHS or Betamax?

  85. Dwave has made the first incremental step towards true quantum computing. The giant leap towards true quantum computing will involve a variation of optical computing utilizing molecularly engineered electro-optic polymers.

  86. Quince

    The point is the people are always looking to put something down. Instead of having an open mind on how it is possible.

    and tomorrow you will see how it is possible and that is the road to the future.

  87. Pingback: Oggi il primo computer quantico ?

  88. Pingback: » Under the radar quantized space

  89. Geordie,

    I was at the demo this morning, cool stuff, congratulations!

    One question I was not able to ask: when do you expect (+- 1 year) you will become fast enough to be competitive with traditional architecture on meaningful problems.

  90. I am a complete novice when it comes to QC but I imagine that if a true AQC has been built then certain government agencies might be a trifle concerned about the ‘national insecurity’ ramifications. I presume that governments will do all in their power to restrict access to QC, particularly once it approaches conventional super computing power.

    Reminds me of the movie Sneakers.

  91. I attended the conference on Quantum Compusing held last year at the Danish Academy of Sciences and arranged by the leading QC experts around the world, sponsored by the Niels Bohr Institute here in Copenhagen. I spoke to most of the experts during the conference, and most of them were saying that a practical quantum computer was 10-20-30 years further down the pipeline!

    Therefore, I find Dwave’s announcement a bit curious, so I hope those of you who can attend the demonstration today will leave us an account!

  92. Pingback: Orin: La primera computadora cuntica de la historia - psicofxp.com

  93. Pingback: randform » Blog Archive » commercial quantum computers

  94. Hey Joep,

    “The point is the people are always looking to put something down. Instead of having an open mind on how it is possible.”

    Don’t let is seem as though these bad people are trying to put anything down. Not about that. Science seperates bad ideas from the good by constantly applying skepticism and trying to blow a given theory out of the water. These are vital signs of a healthy science. Questioning is good, doubting is good.

    If DWaves new technology withstands the scrutiny of the scientific community by providing irrefutable evidence – then it will be a genuine success. And who will be able to attack them then?

    So died cold fussion once upon a time. It isn’t coming back because it fell under the weight of scientific skepticism. But it continues to haunt forums and blogs because of faith. Former is good. Latter is bad.

    “..and tomorrow you will see how it is possible and that is the road to the future.”

    Everyone on this blog wants this. But again – evidence.

  95. Pingback: Estado de Sitio » El Sistema de Computación Cuántica Orion: ¿nueva revolución informática?

  96. Hi Geordie,
    I saw your presentation yesterday at science world. It was really impressive, though at one point you suggested that your chip may be the coldest thing in the universe. Since many of your audience will be science-wise you may want to be carefull about such statements since I think its largely believed that Bose-Einstein condensates are the coldest things – generally several orders of magnetude colder than the dilution refrigerator.

    I very much like your practical approach to things. Especially your error handling, which is a concept that I held for years – just test it! Congratulations!

  97. well all that seems great, but at the same time this is like a science fiction movie, and all this is like matrix or a paralel kind of life, the reality now is not as we know we are in the borderline between the unexpected and amazing :)

  98. “NP complete” is a mathematical notion involving what happens when you try to solve larger and larger problems, *In The Limit* as “N”, the number of bits needed to describe the problem, goes to infinity. Moreover it refers to *Exact* solution of such problems. You have a nifty piece of hardware (in fact we used to call this an analogue computer) which gives an *Approximate* solution when N=16. It is very cool that it is based on the quantum physics of a small number (N) of individual quantum systems in a very nontrival way. Consequently, your reference to “NP – complete” is *Completely Spurious*. But maybe you knew that already and that is part of the cunning marketing ploy?! (People who read what NP-complete actually means, on wikipedia for instance, will realise that they can’t sue you after all, since you can’t possibly have claimed to do anything impossible, at all).

  99. Richard:

    1. NP-completeness is a descriptor of a class of related problems. While the notion of exact solution certainly plays an important role, a problem can be NP-complete and admit approximate solutions which have value to users. In fact nearly all industrial applications of NP-complete/NP-hard problems are solved using heuristics.

    2. Our 16-qubit QC doesn’t give approximate answers. Every problem we’ve run on it gives exact answers, which is of course expected as the problems are small.

    3. The reference to NP-completeness is not spurious as it is the basis of the core philosophy behind the design of the processor. I talk about this in the demo video which is available on youtube and shortly on our website.

    4. I don’t understand your last point, but I can assure you there’s no cunning marketing ploy. Unless telling people what we’re doing is a cunning marketing ploy.

  100. Geordie: “Also just to be clear I don’t believe that the system we’re building is going to exponentially speed up anything. The objective is the quadratic speed up for unstructured search.”

    I really don’t understand quantum computing at all, but I have some ideas about complexity theory. The thing I can’t get my brain around is how a quadratic speed up is going to help much in dealing with NP-complete problems. When the general class of NP-complete problems (code-breaking, for instance…) grows to a real size, quadratic improvements won’t make much difference.

    Put differntly, how much longer keys will protect from your technology ;-)

  101. Well done.
    Adiabatic computation has been alluded to.
    As of the peer reviewed evidence – That must wait until the contracts have been signed – in the same vein as evidence for the weapons of mass destruction wait, proof of the mitigational benefits of nuclear fusion must wait and the rigorous infallibility of the Vista operating system must become mature in order to furnish true wealth.

    In reality, making an analogue computer is NOT the same as harnessing quantum mechanics.
    It would benefit science, to state the difference for the sake of itself.

  102. I have read that in the past vector computers were designed to do one kind of work load (algorithm) really really well. This of course meant other general purpose tasks couldnt be done as well or at all.
    I can almost smell that one fairly generic NP problem is solved really really well in the same way.
    Am i reading this correctly ?

  103. Pingback: TheRoadBlog » Blog Archive » Quantum Computers Becoming Reality?

  104. Pingback: RAD Thoughts

  105. Pingback: Croc’notes » Blog Archive » Infos scientifiques

  106. Hello,
    I came up on a news about the 16 Qbits computer which seems pretty nice.I am nearly out of school (master i guess the rank is in international system) and I hate quantum physics.

    So I was wondering, reading the comments and all I kind of get lost on the gab and everything, so will quantic programmer need an understanding of the physics behind to be able to program for it ?

    Do you know of any How to guide kind of document to quantum programming that would be understandable by someone that is not much into physics?

    Regards,
    Chryso

  107. Pingback: D-Wave Demo « Neil Dickson’s Blog

  108. Pingback: step up transformer

  109. Pingback: Google Wealth System

  110. Im curious for evry today ppl like me how is a 16 Q-bit Quant computer is in Mhz? and how many ” Cores ”?? So i can get a view of the power of Quantum computing =) tkx

  111. Im curious for evry today ppl like me how is a 16 Q-bit Quant computer is in Mhz? and how many ” Core ”?? So i can get a view of the power of Quantom computing =) tkx

  112. Pingback: nextquant Blog Interview with Geordie Rose, D-Wave Systems (Part I) «

  113. Pingback: Lifeboat Foundation Blog » Significant Nanotechnology related developments

  114. Pingback: Moichael

  115. Pingback: current science articles

  116. Hi! Mr.Georde i too much appreaciated with regarding to your invention and i proud with that new invent of this generation.I hope that you give me some way of domenstration and teaching for our class assignment. Iask if how many transistor you used at this year of 2007 and what is the defference with that in DNA COMPUTER?.

  117. Pingback: Software utilities » Desktop monster from Alienware

  118. Pingback: The Axion » Blog Archive » Komputer Quantum Telah Tiba!

  119. Pingback: casino download gratis

  120. Pingback: telecharger partie poker

  121. Pingback: jeu baccarat en ligne gratuites

  122. Pingback: juegos casino ruleta

  123. Pingback: gratis poker im internet

  124. Pingback: www casino proximite

  125. Pingback: play free poker online free online poker

  126. Pingback: payday cash advance washington payday cash advance

  127. Pingback: jouer au poker francais

  128. Pingback: slots spielen gratis

  129. Pingback: poker gratui

  130. 딸국질은 횡경막의 경련에 의해서 일어 나는 것으로 쇼크를

    준다거나 잠시 호흡을 멈추면 낫는다.

    하지만 경우에 따라정신적 부담감으로 빨리 멎지 않을 때가 있다.

    멈추게 하는제일 좋은 방법은 심호흡을 한 뒤 견딜 수 있는데까지

    숨을 쉬지 않는 것이다.

    또는 숨을 멈춘채 찬물을 조금씩 마셔도효과가 있다.

    그래도 가라 앉지 않으면 조용히 숨을 내쉬면

    서 아랫배를 들이밀거나 때때로 배에 힘을 가득 준 뒤 호흡

    을 멈추는 복식호흡도 좋다.

    출처:http://cafe.daum.net/lifebean

  131. hi.. i’am doing a seminar on quantum computers. could you please help me out? its my seminar for final yr of bachelors in engineering. so i’am looking forward for your reply.

  132. hi.. i’am doing a seminar on quantum computers. could you please help me out? its my seminar for final yr of bachelors in engineering. so i’am looking forward for your reply.

  133. Pingback: HCG Levels in Early Pregnancy

  134. Pingback: A detour into the strange world of quantum computing | Ars Technica

  135. Pingback: Il transistor più piccolo del mondo arriva 10 anni prima del previsto | xvid entertainment news

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s