Comment by dm2 | Hacker News Reader

Comment by dm2 | Hacker News Reader

Parent Full thread

Here is another article today bashing the D-Wave Two computers: http://www.wired.com/2013/06/d-wave-quantum-computer-usc/

These machines are testing a completely new computing concept. We're trying to learn how they work so that we can apply that to building machines that can actually outperform traditional computers, and once that's achieved then the potential benefit will be enormous.

According to http://www.dwavesys.com/d-wave-two-system as you scale up qubits the power demands do not increase. That alone has huge potential.

Maybe new materials need to be developed to see the full potential of these computers, lots of research needs to be done. Why are all of these articles so pessimistic?

NASA, Lockheed, and Google purchasing these multi-million dollar machines should indicate that there is at-least some potential and value in the D-Wave computers, even if in the end they just learn what D-Wave did wrong.

https://www.youtube.com/watch?v=CMdHDHEuOUE

"We don't know the best questions to ask the quantum computers, that's what we're trying to find out now."

>We're trying to learn how they work

Is the operation of the D-Wave an unknown? I don't know much about it.

No, we don't have a perfect understanding of what's going on.

The qubit processor is like an intermediate step, instead of a traditional processor, the grunt of the work of special problems is handled by the qubit chip.

The data is put in and the results are as expected, but they are not as fast as modern computers, so now the challenge is figuring out if different types of calculations can increase that performance or what can be done to improve the chip or other methods.

https://www.youtube.com/watch?v=HKUZ6IuJyHw

The machine itself seems so simple though from a visible hardware perspective, most of the D-Wave Two is just to keep the main chip cool and shielded.

It's just experimenting right now. The internet was probably dismissed by many because of it's enormous expense to implement, but the value could be envisioned by many and it paid off.

There is potential for future versions of these quantum computers to be able to solve problems that massive amounts of supercomputers would have trouble solving, and the energy usage is enormously less (same for a small calculation as a huge calculation).

Encryption is one thing that could be hurt by this technology. Instead of millions of years to crack encryption it might be solvable instantly. That's all just what I remember reading about the subject, there's plenty of information out there that's probably more reliable that my babble.

So negative because there's still a non-zero probability that its not going to work at all, ever. That there's something about quantum computing we've figured out wrong, and no benefit is possible.

They're also negative because no one has actually proved that the D-Wave is, in fact, doing any actual quantum computing. That's been the issue all along with D-Wave: no one can prove that it's really doing anything with those qbits.

It depends on your definition of quantum computing, the D-Wave does http://en.wikipedia.org/wiki/Quantum_annealing and has been faster in some applications.

https://plus.google.com/+QuantumAILab/posts

They're just getting started, this field is very new, very difficult, and there are a lot of very smart people trying to answer the what/how/why/where/when of "quantum" computers.

The hardware outperforms off-the-shelf solvers by a large margin

"In an early test we dialed up random instances and pitted the machine against popular of-the-shelf solvers -- Tabu Search, Akmaxsat and CPLEX. At 509 qubits, the machine is about 35,500 times (!) faster than the best of these solvers. (You may have heard about a 3,600-fold speedup earlier, but that was on an older chip with only 439 qubits." [1]

[1] https://plus.google.com/+QuantumAILab/posts/DymNo8DzAYi

> The hardware outperforms off-the-shelf solvers by a large margin

But it doesn't outperform classical algorithms specifically written to solve the types of problems that the D-Wave machine accepts.

Yeah, and I've never seen anyone outside D-Wave show these types of speedups. Last article I read, in fact, was struggling really hard to even show that D-Wave was the same speed as traditional algos.

Just like early digital computers couldn't out-perform specialized mechanical analog computers.

My point is that even if we can't prove the D-Wave Two is faster at any problem than a traditional computer that the massive amount of knowledge gained about quantum computers will outweigh the costs.

It's far too early to worry about a non-zero probability of it not working, just like fusion power. All it takes is one breakthrough and a revolution in computing will begin. Doing research on quantum computers is a low-risk and high-reward situation. None of these companies are betting everything on quantum computers.

Certainly, but maybe it's not a business yet?

Reply on news.ycombinator.com