The model consisted of mathematical representations of amino acids in a lattice, connected by different interaction strengths. The D-Wave computer found the lowest configurations of amino acids and interactions, which corresponds to the most economical folding of the proteins. It worked, but not particularly well. According to the researchers, 10,000 measurements using an 81-qubit version of the experiment gave the correct answer just 13 times. This was owing, in part, to the limitations of the machine itself, and in part to thermal noise that disrupted the computation. It’s also worth pointing that conventional computers could already solve these particular protein folding problems.
This thing can only perform a single type of calculation, and it's terrible at it. Essentially useless.
On December 25 2012 04:22 iNSiPiD1 wrote: The USC, where the first D-Wave computer has been installed, recently published their very first paper which used the D-Wave to perform computations recorded in the paper. I'm not going to pretend that I understand the paper, as I am not a quantum physicist, however I skimmed through it, and it is legitimate. The paper can be found here http://xxx.lanl.gov/pdf/1212.1739.pdf.
The result in Appendix H, could perhaps mark the first bit of evidence that there are some quantum effects going on inside the machine however.
Although there are still some unknowns, there is little doubt in my mind that this company is being run be genius. I don't see any reason why one should expect that quantum computers will, from day one, revolutionize the world. Even if D-Wave is not the future of quantum computing, they have performed a great accomplishment nevertheless.
There's no doubt that there are quantum effects being used in D-Wave's system, we've known that since 2011. D-Wave made a quantum annealing computer, which, while impressive, is not a "pure" quantum computer. Annealing is a technique used commonly to solve optimization problems, which we've typically done using "classical" means. Quantum annealing is similar to another common technique, but a little faster, which is why D-Wave's system is better at certain optimization problems, but that's it.
The reason we're still really far from a true quantum computer is that we haven't figured out how to get lasting entanglement, which is very important in order to make a qubit-based processor actually any faster than a regular one. D-Wave's system exploits a unique loophole in that quantum annealing is faster than thermal annealing regardless of whether there is entanglement or not.
What do you think about section 1.4 - How does quantum mechanics help?
It is true that classical computers literally cannot provide optimal solutions to many problems that contain a lot of information. For example, simple problems such as what is the shortest Hamiltonian circuit among say, 40 or so different cities would take practically forever to compute using current classical computers. Therefore we had to develop algorithms that provide reasonably good solutions, for example.
However to me, it seems as though section 1.4 is implying that they can surpass this problem with their quantum computer. I'm not sure of what the difference between quantum "annealing" versus your "pure" quantum computer, but if it is capable of doing this then it is more than just a "little bit faster" as you claimed. It's turned something that was impossible before into something that is possible.
edit: added the link
D-Wave's system (just like most types of 'real' quantum computers) can be simulated by classical turing machines (classical computers). They will therefore not be able to solve the most relevant class of "difficult" (np-complete) problems (unless P=NP) even though they will certainly push the practicle boundary of computing quite considerably. D-Wave's quantum annealing is further "only" offering a heuristic solution method, so they cannot guarantee optimal solutions. So most of the hyperbole in this thread and on their webpage is way off, even though it is a remarkable piece of engineering.
On December 24 2012 15:06 corpuscle wrote: I'm not even going to try to address the grossly inaccurate statements you make in the OP about what quantum computers are theoretically capable of. D-Wave generally does not have the best reputation inside the scientific community, and there's a lot of prominent researchers/experts in the field who strongly doubt the veracity of a lot of their claims.
Basically, a lot of things are still majorly up in the air about whether they've actually produced a quantum computer or just a highly specialized and very fast "classical" computer. What they do have, to the best of my knowledge, is a system that produces approximate solutions to NP-hard problems faster than a "classical" system for small input sizes, but they haven't truly shown how well it scales for real-world problems, and there are a lot of engineering-related issues that crop up with relatively new technology like this, which hurts actual solve times in the real world when you put it next to a "normal" computer that we understand much better.
Don't get me wrong, what they've made is damn impressive, and appears to be a major step in the right direction, but this isn't some sort of incredible breakthrough that the scientific community is shitting its pants over.
edit: if you're interested in quantum computing, by the way, a much larger breakthrough was recently made by some Australian researchers. It's a little technical, but a quick read-through of that article should hopefully give some of you a better idea of how far we actually are from a true quantum system, and why a lot of people are extremely skeptical about D-Wave's claims.
On December 24 2012 14:40 Zooper31 wrote: Couldn't find a price for the life of me. Only thing that interests me in this thread sadly and it's not available.
The model consisted of mathematical representations of amino acids in a lattice, connected by different interaction strengths. The D-Wave computer found the lowest configurations of amino acids and interactions, which corresponds to the most economical folding of the proteins. It worked, but not particularly well. According to the researchers, 10,000 measurements using an 81-qubit version of the experiment gave the correct answer just 13 times. This was owing, in part, to the limitations of the machine itself, and in part to thermal noise that disrupted the computation. It’s also worth pointing that conventional computers could already solve these particular protein folding problems.
This thing can only perform a single type of calculation, and it's terrible at it. Essentially useless.
The model consisted of mathematical representations of amino acids in a lattice, connected by different interaction strengths. The D-Wave computer found the lowest configurations of amino acids and interactions, which corresponds to the most economical folding of the proteins. It worked, but not particularly well. According to the researchers, 10,000 measurements using an 81-qubit version of the experiment gave the correct answer just 13 times. This was owing, in part, to the limitations of the machine itself, and in part to thermal noise that disrupted the computation. It’s also worth pointing that conventional computers could already solve these particular protein folding problems.
This thing can only perform a single type of calculation, and it's terrible at it. Essentially useless.
Did you even read the article? It said that it's not good at the specific type of problem they threw at it. Science is really fucking complicated and this is a very new technology with a very specific skillset, and they're just trying to figure out what it can and can't do. It's been proven to be reasonably good at other things, which is a pretty big deal, since this is the first computer of its type.
On December 24 2012 14:40 Zooper31 wrote: Couldn't find a price for the life of me. Only thing that interests me in this thread sadly and it's not available.
The model consisted of mathematical representations of amino acids in a lattice, connected by different interaction strengths. The D-Wave computer found the lowest configurations of amino acids and interactions, which corresponds to the most economical folding of the proteins. It worked, but not particularly well. According to the researchers, 10,000 measurements using an 81-qubit version of the experiment gave the correct answer just 13 times. This was owing, in part, to the limitations of the machine itself, and in part to thermal noise that disrupted the computation. It’s also worth pointing that conventional computers could already solve these particular protein folding problems.
This thing can only perform a single type of calculation, and it's terrible at it. Essentially useless.
Thank you, that's hilarious.
so, it only does one thing and my computer can already do it? And it costs $10mil? lol
The model consisted of mathematical representations of amino acids in a lattice, connected by different interaction strengths. The D-Wave computer found the lowest configurations of amino acids and interactions, which corresponds to the most economical folding of the proteins. It worked, but not particularly well. According to the researchers, 10,000 measurements using an 81-qubit version of the experiment gave the correct answer just 13 times. This was owing, in part, to the limitations of the machine itself, and in part to thermal noise that disrupted the computation. It’s also worth pointing that conventional computers could already solve these particular protein folding problems.
This thing can only perform a single type of calculation, and it's terrible at it. Essentially useless.
Did you even read the article? It said that it's not good at the specific type of problem they threw at it. Science is really fucking complicated and this is a very new technology with a very specific skillset, and they're just trying to figure out what it can and can't do. It's been proven to be reasonably good at other things, which is a pretty big deal, since this is the first computer of its type.
I understand that quantum decoherence is the biggest problem for quantum computing. Can someone with more insight explain what is the current status of QEC and if this finding is a good step forward?