It is not needed to find “the” solution (but it would be nice). The point is that now we can treat given protein with enough accuracy for the problem of molecular mechanics.

So, if our most powerful classical computers have a hard time solving NP-complete problems, what is it about the brain that is giving it an advantage? Is is purely a case of parallel processing? If so, why can’t we design a computer than can do parallel processing like a brain can? Is it a problem of implementation?

Also does getting a million gamers to challenge something in general not count as brute force?

That being said, I think it’s brilliant that this has been done. Imagine what we could have achieved if this was done by everyone in America instead of them watching tv. It would have been solved before the commercials.

*NP-complete is also considered NP-complete because it is proven that every other problem in the set of NP-complete problems can be transformed into this problem in polynomial time.

On the folding problem, I didn’t see enough information to determine if the solution was *a* solution or “the” solution — I.E., was the entire state space searched to determine that no better solution exists, or did they just find one solution that was sufficient?

If the latter, then it would be equivalent to someone finding a fast traversal in the Traveling Salesman problem (which most people can do pretty well) and declaring the problem solved. While a fast solution is usually good enough for government work, and may indeed be *the* best solution, SOLVING the problem means showing that there is no better solution, and that’s not what you did.

This is, fact, called an NP-problem (non-deterministic polynomial).

Everywhere in this article where you use the phrase “NP-problems”, the more accurate term is “NP-complete problems.” Just saying that a problem is “NP” doesn’t say that it is difficult. “NP-complete” means that the problem is one of the hardest problems in NP, and these are the tough ones. Finding the protein folding structure (when pseudoknots are allowed) that has the minimum free energy is NP-complete.

Such problems are impossible to solve by computational brute force

Believed to be impossible to solve significantly faster than brute force. And as you say, the number of possibilities one must explore with brute force quickly becomes astronomical.

However, some think that quantum computers may be designed to solve NP-problems

This is a common misconception about quantum computing, but it is very unlikely that quantum computers are a magic bullet that make every NP-complete problem tractable. That would require an exponential speedup across the board, while quantum computing only gives a general quadratic speedup. There are some specific problems (like factoring) for which the speedup may be exponential, but these problems are somewhat incomparable in difficulty to NP-complete problems.

And that is probably more than you cared to know about the current state of computational complexity theory Cheers!

However, some think that quantum computers may be designed to solve NP-problems (even if they are not particularly suited to running Windows).

That is probably not the case. Quantum computers can solve a class of computational problems known as “bounded quantum polynomial”. The best-known quantum algorithm is Shor’s factoring algorithm, which shows a dramatic speed-up from classical algorithms. However, factoring a composite number into prime factors is not in the same class of problem as the traveling salesman problem, which is in a class known as NP-Hard. There is no evidence that a quantum computer can efficiently solve NP-hard problems, which is too bad as there are many interesting problems in that class.