Quantum Computing: Superpositon of states problem

[joelio36]

Hey, I'm learning about quantum computing for a project and I'm a bit stumped about a concept:

They say in quantum computing you can have the superpostion of all possible states, then perform an operation on that wavefunction, and thus have all possible states processed in one operation.

That sits fine with me, I get that, but what stumps me is how do you extract all that information without collapsing the wavefunction? Yes you have a superpostion of all the states you need, but as soon as you extract (read: observe) a state, don't all the others in the superpostion disappear?

By the way, I'm writing this to explain quantum computing to my peers, 3rd year Physics B.Sc students, so we aren't the brightest bunch!

Thanks

 

[joelio36]

Bump. Any chance of an input? I can't find anything online. Cheers.

 

[joelio36]

Still nothing hey...

 

[JK423]

Yes, at the end you do measure the wavefunction and all the other states in the superposition disappear. But this measurement gives you the answer you want, so you don't care about the rest.
Remember, with quantum computing you want an answer to a specific problem like in classical computing. The last measurement that you do, which collapses the wavefunction, gives you this answer.
That's all.

 

[Wolfgang2b]

I guess this is one of the reasons why constructing quantum algorithms is tough. It has to be clever enough to extract the information from the superposition.

Algorithms/operations might also collapse the wave function to the particular state that we desire with higher probability but there might be a chance of getting some other output. Then based on the probabilities we have to repeat the computation till we are satisfied.

 

[atyy]

Also, since you collapse the wave function at the end to get the answer, the answer is usually probabilistic like in Grover's algorithm, so you have to run it a couple of times (collapse the wave function of several systems) to get the right answer with high probability.