Enough already
Break the sticker that says "warranty void if broken" and see what's in it. X-ray the CPU and slice it up. I'm sure a lot of people will donate money for the cost of the machine; most notably their competition.
First, the world thought that D-Wave hadn't built a quantum computer; then, it thought there was a quantum computer in the box; next, there was disappointment that the D-Wave machine didn't speed things up (but might still be quantum); and now, it starts to look like it's not quantum after all. In the latest instalment in the …
Break the sticker that says "warranty void if broken" and see what's in it. X-ray the CPU and slice it up. I'm sure a lot of people will donate money for the cost of the machine; most notably their competition.
Actually looking at it, that may be a feasible thing to do. It seems to be a "normal" microchip which you can "easily" reverse-engineer.
http://en.wikipedia.org/wiki/File:DWave_128chip.jpg
... because it's a superposition of a quantum computer and a classical computer. D-Wave are clever.
It's depressing that always I arrive far too late in comments to do the Schrödinger's cat jokes. On the other hand, I only find out when I read the comments, so was I ALWAYS too late to do so?
I'll ask Wigner - he's a friend of mine.
Technically, you cannot know with any certainty whether or not you will arrive too late to post a quantum physics gag on time. Only when you click "Post a Comment" will you cause the wave function to collapse and leave you in a too late or just in time to post your joke state, as clicking this button triggers the observation inside the box which contains the information as to whether you got your joke in on time.
However in the "many worlds interpretation", there may be an infinite number of parallel universes where you always get to post first, and thus gain all the early mod points.
So if you read the data output (ie measure it) you don't know where it is , but if you know where it is you can't read the data?
(Alonso) Church's thesis was that all computing machnices are essentially the same, (but he did not foresee computing based on quantum superposition as far as I am aware).
The reason it is referred to as "Church's Thesis" is that it seems impossible (in the mathematical logic sense) to prove that all (classical) computers are essentially the same (i.e. have the same theoretical computational power, implementations differing by being bounded by things such as memory, processor speed, efficiency of compilation etc.)
There is therefore a (mathematical) logical difficulty in determining whether there may be a model for a classical computation device which would behve like the D-Wave processor meaning it is in essence a very fast Turing machnie.
Quantum computers are not more powerful than classical computers in the Turing sense. They are just faster. Church's thesis is still going strong - to break out of it, you need to muck around with black holes or the like.
The reason it is referred to as "Church's Thesis" is that it seems impossible (in the mathematical logic sense) to prove that all (classical) computers are essentially the same (i.e. have the same theoretical computational power, implementations differing by being bounded by things such as memory, processor speed, efficiency of compilation etc.)
Er... No.
When Church / Turing say that all Turing-complete computational models (not "computers", that's something else entirely) are equivalent, they mean that in a theoretical sense – namely that, given infinite time and resources, all such models can / can't achieve the same things. In that sense quantum computing is no exception – you cannot, for example, compute a solution to the halting problem using a quantum computer.
Then there is the caveat that even this theoretical equivalence (which concerns Turing-complete models alone, remember) is only valid under the condition of infinite time and resources. This is an important restriction: there are many things that can be done with a Turing machine in theory (such as proving a theorem by enumerating and checking all possible proofs) that in real life would take too long, our require too much storage, to be practical.
Only when we take into consideration real-world limitations – time, space, viable implementations – is that Turing-complete models differ on what they can achieve, and then of course some tasks will be feasible for some models / implementations but not for others. In the theoretical sense a trained clerk with paper and pen is as good as a supercomputer, but what one can do in hours the other won't accomplish in their entire life – but that's a practical difference, not a theoretical one. Give the clerk unlimited years and unlimited stationery, and there's nothing a supercomputer can do that they can't.
Then there is the caveat that even this theoretical equivalence (which concerns Turing-complete models alone, remember) is only valid under the condition of infinite time and resources.
Indeed, if all you have are finite resources, the set of functions you can compute can be implemented using a DFA. Just use a DFA with one state for every state you have resources to represent. A mechanical implementation of a TM that's bound by finite resources is not strictly more powerful than a (hypothetical, more-resource-expensive) DFA, even though an ideal TM is strictly more powerful than a DFA.
It's easy to see this in practice. Besides the construction argument above, consider a language such as a^{n}b^{n}, which cannot be recognized by a DFA but can be by a PDA or TM (or equivalent, such as an nPDA or Post machine). If your physical TM implementation has finite resources, you can only recognize words in that language up to a certain finite length. If the length is bounded, you can construct a DFA to recognize that subset of the language; it only exceeds the capabilities of a DFA by being unbounded.
might not.
The more I read about this thing the less like anything I recognize as a computer.
Is it a scam or have the developers fooled themselves?
Who knows?
Do you call it a scam coming a group of real scientists and engineers with backing from Google, NASA, CIA, private investors and governments? They made a completely radical processor competitive with todays CPUs developed over decades with trillions of dollars (for crying out loud). This technology is completely new territory, so I'll cut them some slack.
Here's a box of magnets. Shake it around a bit. You can read tea leaves, can't you?!
So D-wave are not bothering to apply for any patents?
Or are they 'secret patents'? LOL
They have 100 US patents on this thing, check it out:
http://www.dwavesys.com/en/pressreleases.html#dwaveus_100_Patents
I hooked my D-Wave up to my E-Cat and now I am the secret emperor of Earth. Take that, skeptics!
Now to sort out everyone's problems, as soon as I finish this bottle of Dr Snakeoil's Mystery Tonic.
If you have been following the Dwave saga closely you have no idea how much information have been published out there about the machine and it's internal workings. It's no secret and the scientific community knows that, the issue is that it's really hard to measure this type of device even with the scientists and Dwave. They even have a problem agreeing with benchmarks. Please go ahead and check this out so we can have a better informed discussion.
" the company keeps its internals a closely-guarded secret" Yet another black box... whoopee!