D-Wave sells first quantum computer

D-Wave sells first quantum computer

SOURCE

"Canada's D-Wave Systems has won a ringing endorsement of its controversial quantum computer - an order from Lockheed Martin.

For years, D-Wave has claimed to have developed a functioning quantum computer, and for years its claims have been challenged.
[...]
However, a recent paper in Nature appeared to back up its claims - and seems to have been enough to convince Lockheed Martin. It's signed a multi-year contract for the system, maintenance and support, and plans to use it for some of its most challenging computation problems, says D-Wave."

Related News Links:
Technical Specs

Please do not create minimal posts to start your new thread. Please provide your own opinions when posting links or videos.
Starting a New Thread ... Please Look Here First.

I have done a bit of reading on this system, and have learned more about how it works and the controversy behind it. It isn't really like a desktop computer that would be for everyday use, as some might suspect, and the pricetag is currently $10 M. Interestingly, here are some quotes I pulled from Wikipedia:

Umesh Vazirani, a professor at UC Berkeley and one of the founders of quantum complexity theory, made the following criticism:[24]

"Their claimed speedup over classical algorithms appears to be based on a misunderstanding of a paper my colleagues van Dam, Mosca and I wrote on "The power of adiabatic quantum computing." That speed up unfortunately does not hold in the setting at hand, and therefore D-Wave's "quantum computer" even if it turns out to be a true quantum computer, and even if it can be scaled to thousands of qubits, would likely not be more powerful than a cell phone."

Wim van Dam, a professor at UC Santa Barbara, summarized the current scientific community consensus in the journal Nature:[25]

"At the moment it is impossible to say if D-Wave's quantum computer is intrinsically equivalent to a classical computer or not. So until more is known about their error rates, caveat emptor is the least one can say."

Source

So whether LM has made a mistake in this business deal...we will have to wait and see what happens.

According to Scott Aaronson, a Computer Science professor at MIT who specializes in the theory of quantum computing said;

D-Wave's demonstration did not prove anything about the workings of the computer. He claimed a useful quantum computer would require a huge breakthrough in physics, which has not been published or shared with the physics community.

I wonder if there are some quantum physics documentation that hasn't yet been viewed to the public. And if so I hope who ever discovered such breakthrough gets recognition for such great achievement.

Well, whether the D-wave computer works or not, we shall soon see, I suppose. If Lockheed Martin is willing to invest in it, then it's either the real deal or a very clever con, because a company with a reputation like that wouldn't put their money up unless they were confident in the technology.

reply to post by millicake


Thanks, OP. I have watched the video and it has spun me off in to a whole new area of research. Very important stuff; what is spoken about in the video. From what I gather from the guy's talk, these computers will allow us to model the systems that drive the invisible world behind matter; and if we can model it we can build it. Am I on the right track?

The video has led me to discover membrane computing which is really cool because I am really in to Biology right now.

Originally posted by Frater210
reply to post by millicake

 

Thanks, OP. I have watched the video and it has spun me off in to a whole new area of research. Very important stuff; what is spoken about in the video. From what I gather from the guy's talk, these computers will allow us to model the systems that drive the invisible world behind matter; and if we can model it we can build it. Am I on the right track?

The video has led me to discover membrane computing which is really cool because I am really in to Biology right now.

If this is true then you guys do realize that the world would end as we know it. If we can understand it we can build it.........

So what could be more sustainable on this planet than the eradication of the human species, bit we wouldn't know it because we could take a "photograph" of the current reality then build a self sustaining quantum computing/simulating machine that uses robotics to repair and maintain. In this machine the "photograph" would be placed then the machine would be turned on. Based in the location, position, rotation, velocities etc, the universe/reality could be simulated but based on this we would not notice the difference. This would leave the earth to continue on and heal itself from the damage that "reasoning" animals have done to it. It's kind like the matrix theory but more logical, what would be the point of humans as batteries when sun, wind, and geological forces could be used to provide energy.

The only flaw would be figuring a way to avoid complete planetary destruction.

As I typed that I came up with a method....

We could create a self replicating system that would build itself l, copy the reality program (repeatedly), then launch it toward many bare weatherless planets where it could use solar energy to continue the process. The system would also need to he able to self harvest metallic minerals as well. Very complex system but completely plausible.

Originally posted by Frater210


what is spoken about in the video. From what I gather from the guy's talk, these computers will allow us to model the systems that drive the invisible world behind matter; and if we can model it we can build it. Am I on the right track?

The exact use for the system itself has not been publicly addressed. From my understanding, is that their primary focus is the development and intergration of a machine learning application/s.

SOURCE

Machine learning, a branch of artificial intelligence, is a scientific discipline concerned with the design and development of algorithms that allow computers to evolve behaviors based on empirical data, such as from sensor data or databases.
[....]
. A major focus of machine learning research is to automatically learn to recognize complex patterns and make intelligent decisions based on data

Originally posted by Frater210
Very important stuff;

Indeed.

Originally posted by Frater210
Thanks, OP. I have watched the video and it has spun me off in to a whole new area of research.

Your Welcome.
BEST WISHES!

In other words, Lockheed Martin will act as a sort of testing ground for D-Wave, perhaps making the whole deal more of a joint development partnership, rather than a straightforward sale.

So, Lockheed Martin is guinea pig?

I still don't understand what this computer can do, and now I don't feel so bad, because from the sound of that, I'm not the only one. It sounds to me like they want to figure out what it can do.

I think the question is; why does Lockheed Martin want a computer that can learn and develop itself in to AI?

Don't they do airplanes?

Originally posted by Frater210
I think the question is; why does Lockheed Martin want a computer that can learn and develop itself in to AI?

Don't they do airplanes?

Well, we are in the process of taking pilots out of the airplanes, right?

One of the biggest constraints on the maneuverability of advanced craft is the ability of the pilot to withstand G-forces, so one way to solve that is to not have a human pilot in the plane. A pilot can fly it remotely but I imagine you don't have quite the same perspective behind the joystick as you would from the cockpit, so having some AI on board to replace the on-board pilot might be desirable, if a little "Terminator-ish".

Originally posted by Frater210
I think the question is; why does Lockheed Martin want a computer that can learn and develop itself in to AI?

Don't they do airplanes?

Stability control, pilot assist, and target recognition.

We did some electrooptic target recognizers for LM and Boeing, at least the electronic bits of them, not so much the holography parts, and they were at least moderately kickass. They could be 'trained' to some extent by showing them images of targets in various aspects, then they could learn to perceive the targets when hidden by camo or foliage or whatnot. Really sweet application of physics.

ETA: there are target recognition algorithms that depend on qc that hasn't been invented yet - most of them use Grover's Algorithm for recognizing an object image from a superposed image of all possible orientations and aspects, I would guess you could use QSA to do the same thing. The current SOTA involves using holographics to do the thing. As a practical twist, I'd guess you could one day expand this to machine vision so that you could actually get real-time machine perception of real-world objects without all that pesky serialized edge detection crap.

Besides which, you ought to try piloting an inertially decoupled intersystem FTL craft - the control state space is all poles and zeroes, moving all the time. Doing it in DSP type processors is barely possible - with a quantum AI buffering the drive controls and helm, you might get more of them home. And you know LM and Boeing are the ones building them, right?

Originally posted by Arbitrageur

In other words, Lockheed Martin will act as a sort of testing ground for D-Wave, perhaps making the whole deal more of a joint development partnership, rather than a straightforward sale.

So, Lockheed Martin is guinea pig?

I still don't understand what this computer can do, and now I don't feel so bad, because from the sound of that, I'm not the only one. It sounds to me like they want to figure out what it can do.

I continued to research more on patents under D-wave INC. and what I found was that they're the leading producers of Quantum processors. There's a high probability that Lockhead Martin are an acting roll on placing the pieces together, kind of like the computer geek that customize their own computer by buying different parts and then building it (the computer). That would explain this quote:

perhaps making the whole deal more of a joint development partnership,

Systems, devices, and methods for controllably coupling qubits

Quantum processor

Architecture for local programming of quantum processor elements using latching qubits


D-Wave corporation is one of many players in the game. A scalable quantum computer has successfully been patent.
Scalable quantum computer

Google has also worked with DWave quite a bit the last year or so.

They have used it to improve their image search, and for Google goggles - which is basically a mobile app that lets the user take a picture with thier phone to get search results. They seem to have gotten positive results from the system, so I would not be surprised if Google becomes their next customer.

reply to post by Bedlam


Dude, I won't repost your whole post but I hope that others will pay attention to it. Marvelous amount of info in a small space. Thanks for the heads up it makes perfect sense. may I humbly offer my thread here...

www.abovetopsecret.com...

It is about Ethical Adaptation software and the OS's that run war robots. I would really like to understand them more deeply and would be really interested in what you think.

Also I am wondering a lot about what it means that they are using computers used to model biological processes like membrane computing.

Thanks.

Originally posted by Frater210
Also I am wondering a lot about what it means that they are using computers used to model biological processes like membrane computing.

Membrane computing is one of those things I'm only marginally familiar with. It's only started popping out in the last few years, and my vague and possibly incorrect take on it is that it's one of those computational forms that doesn't have a native way to DO it, that is, there's no membrane processor per se, so any membrane computation you do is by its nature a simulation of what a membrane processor would do. Only doing that with a von Neumann or Harvard machine is going to be slow as the nature of membrane processing is that it happens all at once and it's massively parallel. For that sort of thing qc is likely to be more useful, if you can apply it. QC is not for general purpose problems, it's sort of like using holographic computing in that regard.

Long ago and far away we worked on an offshoot of the EO holographic targeting system that would do massively parallel optical computation in a neat little package about the size of a shoebox. The sorts of things it was good at didn't lend itself to generalized desktop type usage, you couldn't use it to get your frame rates up in Crysis, for example, although it was hell on wheels for some very specialized number munching. I know they built some of them for someone, and about ten years after the fact, an Israeli company "invented it" and was marketing the thing. Only their invention was the same unit, down to the sprue marks on the laser unit casing.

reply to post by Bedlam


Thanks for all of that. It really is supremely interesting stuff.

reply to post by millicake


Now this should have everybody rolling.

I wonder what Clif High would say about this this little invention he failed to predict.

Since this is real, he should shell out 10 million for one of those bad boys and make some real predictions!

Oh wait.... probably wouldn't work for that. lol .


HaHaHA