Quantum Computing Explained! (Well, Sorta)

An anonymous reader writes "Valiant effort to 'explain' quantum computing over on silicon.com — covering the difference between classical computers and quantum machines."

One minor mistake

[vlm]

One more thing, there is a minority of scientists who believe that building a quantum computer will turn out to be out-and-out impossible.

However, if those scientists are right, the implication of not being able to build such a machine is that quantum mechanics itself, as a description of nature, is wrong. Either way, the stakes could not be higher.

One possible failure mode is the theoretical power required could exceed the light fluxs of the visible universe, that would be a bummer. Maybe in true supercomputer style, a formerly computational problem is merely converted into an I/O problem, the interface to the classical world might be too slow/imprecise/analog/noisy/random to pull useful results out of it. Nothing wrong with quantum theory at all, just not possible to interface usefully with the greater classical world.

Or the more practical engineering/accounting failure mode where it would simply be cheaper / faster / more efficient to use mass produced classical processor, possibly for any problem.

Quantum Co-Processor?

[Anonymous Coward]

Everyone always talks about the differences between a standard computer and a Quantum computer. Graphics cards are good for floating point numbers, why can't we have a Quantum Card to handle quantum operations? Does it really have to be one or the other?

I know you're being funny but

[Tekfactory]

Do you remember the Google Quantum Powered Image Search
http://www.newscientist.com/article/dn18272-google-demonstrates-quantum-computer-image-search.html [newscientist.com]

Some folks have questions about D-Waves technology, but there are people at Google who have been writing applications for Quantium computers.

Re:One major mistake

[Bill, Shooter of Bul]

All theories are really just models of the universe, some work better and tell us more about how the universe does work. Quantum mechanics does tell us many very important things about our universe. Most importantly, and confusingly for everyone first learning it, our universe is not deterministic. Unlike dice, where we could predict with absolute certainty what the outcome would be if we collected enough data about the throw, we cannot do that in the quantum world. There are no "hidden variables" that we could use to increase the accuracy of our probabilistic predictions.

See bells theorem [wikipedia.org] for more mind bending details.

Re:One major mistake

[Anonymous Coward]

Note that not everyone rejects hidden variables. Claiming that QM implies a non-deterministic universe because of the absence of hidden variables is in fact subtly wrong. The dominant interpretation of QM (which claims the nonexistence of hidden variables) *assumes* nondeterminism, it doesn't conclude it. You can find a complete quote on the Wikipedia page for Superdeterminism, but there was an assumption in the design of the EPR experiment that assumed non-determinism as a means of preserving the free will of the experimenters. In other words they mixed philosophy into science then got freaky weird results, and now most of their followers fail to consider the possibility of a causal link between the two.

Re:Horrible

[Interoperable]

Let me take a run at explaining quantum computing less awkwardly than the article.

A quantum bit (qbit) may be in a 0 state, a 1 state or any linear superposition (combination) of the two, eg. 0 + i1. When measured, the outcome of the measurement can only be 0 or 1 with the probabilities of each being governed by the ratio of contributions to the qbit from the 0 and 1 components.

One qbit can usefully encode one bit of classical information (this is the point that most articles on the subject muddle up). Entangled qbits, however, also encode information into the relationship between them. More accurately, the state of the two qbits combined cannot be described by the individual states of each qbit. The number of possible states that the combined system can occupy is greater than the number of states that two unentangled qbits (or classical bits) could occupy. In other words, N entangled qbits occupy a much larger state space than 2^N, which is the state space for N classical bits.

A quantum register containing N qbits can yield an answer with, at most, 2^N bits of classical information once measured; however, the computation itself can be performed in a state space that is much larger than 2^N, hence the dramatic increase in computational power for certain algorithms. It's difficult to come up with algorithms that exploit the large quantum state space but yield a deterministic (rather than probabilistic answer). In some cases, however, even probabilistic answers may be okay if the correctness of the solution can be verified quickly with a classical algorithm and the quantum computation re-run.