A Look At Quantum Computer Manufacturer D-Wave and Its Founder

First time accepted submitter tpjunkie writes "Many slashdot readers will remember D-wave's announcement in 2007 of its quantum computer, an announcement met with skepticism and a good amount of scorn. However, today the company has sold quantum computers to such companies as Lockheed Martin and Google, and their computers have gone from a handful of qubits to 512 in their most recent offerings. Nature has a story including an interview with the company's founder Geordi Rose, and a look at where the company is headed and some of the difficulties it has overcome."

Comments

Not a QC!

[AchilleTalon]

The summary is saying it is a quantum computer because it sold these to Lockheed Martin and Google. Please. stop that shit. They are pretty fast computers, however nobody has proven it is quantum computers. Even the CTO at D-Wave is not able to demonstrate it and he just doesn't care saying it is damn fast and that's all matter for him.

Slashdot should stop advertising D-Wave computers as QC until it has been proven.

• http://www.npr.org/2013/05/22/185532608/quantum-or-not-new-supercomputer-is-certainly-something-else
  ”What we do is build computers,” Rose says, “and if we can build the fastest computers the world has ever known, you can call them whatever you like, and I’ll be happy.”
• http://www.scottaaronson.com/blog/?p=1400
  "Instead, journalists have preferred a paper released this week by Catherine McGeoch and Cong Wang, which reports that quantum annealing running on the D-Wave machine outperformed the CPLEX optimization package running on a classical computer by a factor of ~3600, on Ising spin problems involving 439 bits. Wow! That sounds awesome! But before rushing to press, let’s pause to ask ourselves: how can we reconcile this with the USC group’s result of no speedup?"

Re:Not a QC!

[tpjunkie]

I submitted the article. I called it a QC, because if you read TFS, there are a couple of papers linked indicating that there seems to be evidence that the machine is functioning as an adiabatic quantum computer. Of course, these results have been challenged. However, for the purposes of a summary, it seemed in my mind, ok to call it what the manufacturer does, which is an adiabatic quantum computer.

Re:Not a QC!

[Arkh89]

Let's make every one happy :
D-Wave = \frac{1}{\sqrt{2}} \Psi_{classical computer} + \frac{1}{\sqrt{2}} \Psi_{quantum computer}
But, PLEASE, don't measure it, seriously...

Re:Not a QC!

[Anonymous Coward]

A computer directly using quantum effects is a quantum computer.

Transistors directly use quantum effects to work, yet we don't call desktop computers "quantum computers".

Most of the interest around "quantum computers" exists because there's good reason to believe that they can yield improvements in the asymptotic time complexity in solving some problems (the mathematical definition would be something like "can solve BQP problems efficiently"). Before Shor's algorithm, almost no one cared about quantum computing.

So, if all you have is a special-purpose computer like D-Wave's, which just uses some quantum effects to calculate something, but you're not sure it can solve BQP problems efficiently, it's debatable whether it should be called "quantum computer".

Re:Not a QC!

[slew]

Transistors directly use quantum effects to work, yet we don't call desktop computers "quantum computers".

The transistors in the CPU in your desktop computer are IGFETs (insulated gate field effect transistors). The principle of operation of this device is that moving charge on the gate can enhance or deplete the number of mobile electrons in the source to drain channel under the gate and cause it to turn on or off using an electric field effect which is not considered a quantum effect.

To be fair, at the scales that modern transistors operate, there are some interesting quantum effects. Most are considered as "bad" (causing problems with the "classical" operation of the transistor by tunneling charge or changing thresholds), but there are a few things like strained silicon that are used to improve performance (which used to create quantum containment and effective mass modifications to make small geometry operation more feasible), but these quantum effects aren't intrinsic to the operation of a generic IGFET (just a FET that's really small).

There are of course stuff in your desktop computer that intrinsically rely on quantum effects to work. For example, the flash memory (uses tunneling to move charge in and out of an isolated control gate). However, there are many other things that are similar to the transistor's use of QM effects like the optical drive (solid state laser uses bandgaps to get a certain frequency) and the disc drive (uses the GMR effect which is related to QM electron spin transport), but that's really just to make stuff work when it is really small, not intrinsic to the operation.

In the end, it's all physics and computers use physics and when you make things really small the quantum nature of physics must be accounted for, but it can be taken advantage of too. As for calling a quantum adiabatic computer a "quantum computer" I agree that would be a no. It technically relies on tunneling, so it's sort of like a flash memory in that respect (it's basic theory of operation requires a non-classical QM effect which is different than a transistor).

As to whether D-Wave actually does or doesn't implement a QM adiabatic algorithm, or perhaps just uses QM tunneling to improve a more classical annealing implementation speed and result, and if that actually makes any practical difference, is another question.

Re:Not a QC!

[sFurbo]

Transistors directly use quantum effects to work, yet we don't call desktop computers "quantum computers".

The transistors in the CPU in your desktop computer are IGFETs (insulated gate field effect transistors). The principle of operation of this device is that moving charge on the gate can enhance or deplete the number of mobile electrons in the source to drain channel under the gate and cause it to turn on or off using an electric field effect which is not considered a quantum effect.

I think the point is that a transistor does this through a quantum effects. It works due to the energy levels of the semiconductor, which are definitely a quantum mechanical in nature. You could do the same with a vacuum tube, so there is nothing inherent in the effect of a transistor that is quantum mechanical, but the the level that flash memory is quantum mechanical (the effect can be obtained differently, but in this case it isn't), so is a transistor.

not to sound picky

[nimbius]

im not sure how best to phrase this, but its not a quantum computer in the absolute sense. Its more of a computer in a quantum state that acts as an annealer. all it does is find the global minimum of a given objective function over a given set of candidate solutions. companies that buy it should at least be given full disclosure that its basically a ten million dollar math co-processor...one where depending upon the solver and the equation, mileage may seriously vary. traditional computing has been conjectured to be, at the cost of the D-Wave, not only faster but cheaper.

funny thing

[wbr1]

With quantum computers you can tell if they exist or if they work but not both. The moment you determine both it becomes a regular computer. Or a brick.

Re:the missing fine print

[SirSlud]

I guess D-Wave's Google Ads campaign (Buy a 10 million dollar computer! Click here!) wasn't getting good sell through rates, huh?

Re:quantum computing

[hawguy]

This technology won't be impressive until it can perform general computing tasks. Right now, it's too constrained of a technology to be useful for something as simple as web browsing. Great promise... but that's what it is: A promise.

There are lots of specialized computers that do one thing really well, yet still aren't great at general computing tasks... like GPUs and DSPs. For that matter, the CPU in your hard drive may not be nearly powerful enough to run a web browser, yet it's still extremely useful for its intended purpose.

Not every computer development is meant to make Firefox run faster.

I'll believe it when I see it

[Myria]

Wake me when someone makes a 2048-qubit quantum computer that can run Shor's algorithm. The Xbox public key and I have some unfinished business.

Re:quantum computing

[spyke252]

The research into quantum computing is using done with the goal of a universal quantum Turing machine, which would, by proof, run classical algorithms in addition to quantum ones.

Not the D-Wave. There's two branches in current quantum computation: General quantum computation, which is still stuck at the implementation stage (of which languages like QCL derive) and D-Wave's computation (which, admittedly, is geared toward quantum annealing and no other quantum procedures, and is therefore not a general quantum computer).

If I were to think a few years down the road, the path D-Wave is taking would culminate in chips that do specific things, such as perform quantum communication protocols, but only those things that were hardwired into the chip. It's hard to think of how a quantum operating system or a quantum programming language would operate under such a model. The general quantum computing path, for which four major quantum programming languages have been written already (QCL, LANQ, CQPL, and QML), if possible, would allow for Turing-Complete machines.

Re:quantum computing

[arth1]

There are lots of specialized computers that do one thing really well, yet still aren't great at general computing tasks... like GPUs and DSPs.

And even analogue computers, giving a near immediate result where a digital or GP computer would have to do complex calculations.

Re:quantum computing

[manicb]

I've used a BlueGene/Q. It sucks for web browsing and gaming, the processors are only 1.6 GHz and it doesn't even run Windows ;-)

Implementing MPI scalable parallelism into Firefox would actually be a pretty interesting exercise for somebody with the time on their hands...

It is not fast either

[amaurea]

The one thing the D-Wave computer is good at is solving the "D-Wave problem", or things that can be expressed in terms of that problem. However, even at this, its speciality, it is 12000 times slower than a normal single-core computer. The reason why some were reporting that the D-wave computer was faster than classical computers at this problem was simply that they used a very inefficient program to do this.
http://www.archduke.org/stuff/d-wave-comment-on-comparison-with-classical-computers/

So basically: There might be some interesting things going on with the D-wave computer, but none of these are practically useful. As far as I understand it, if somebody replaced the insides of a D-wave computer with off-the shelf computer parts, then you wouldn't be able to tell the difference, except it would be much too fast unless artificially slowed down. That makes it hard to see why anybody would pay millions for a computer like that. It also doesn't help that D-wave is completely opaque about the science and engineering behind the computer.

Re:Use for Bitcoin mining?

[gl4ss]

I'm surprised no one has suggested it yet...

d-wave can't be used for bitcoin mining. or can, but extremely slow.

a real quantum computer could be.

Re:Guts & Determination

[gl4ss]

How does selling to Google & Lockheed constitute sucking, since they evaluated the DWave and then bought it in a free market transaction.

well the fraud portion is that they thought the machine was any good for a problem they have and not beaten by desktop machinery in what it is supposed to do best...

lockheed billed it to government projects. neither company is willing to admit that they spend millions on shit - or are hoping the next gazillion million version will actually do something. adding units doesn't help at all either since that doesn't turn it into a real quantum computer.