So exactly how will quantum computing change software, hardware and our IT society in general?
Like, how will quantum computers be better than current computers in terms of how advanced the software can be, and what applications will it have? Sorry if my question is retarded, obviously I'm not much of a /sci/entist
Ive seen this question alot and every time the answer has been, 'they wont be better in general, you'll never play crysis on a quantum computer, they'll be good for some very specific kinds of computing tasks'
I dont really understand why, but thats come up every time ive seen it
>>7657787
Theoretically quantum computers would be able to utilize non-deterministic algorithms to allow infinitely-parallel processing in order to solve an algorithm. You give it an algorithm (again, this is all still theoretical) and it will solve that algorithm with every infinite possible range of variables and give you only the range of variables that were able to pass through the algorithm.
While this kind of infinitely-parallel processing sounds like it would be wonderful for all the parallel graphics rendering in the like, remember that when rendering graphics and processing mass data, you're not simply performing the same tasks a million times over a large sample set, but you want a million different answers out of it. Quantum Computers can only give you one answer at a time.
That's not to say it would be completely useless. For one, it'd be a decryptor's dream. But the entire field of computer science is built upon deterministic computation and the realm of non-deterministic computation remains relatively unexplored except by a few mathematicians.
>>7657806
Exactly. After Shor proposed his algorithm, there has been a flurry of papers how to use quantum computers for quantum chemistry. There is simply not enough concentrated effort on quantum algorithms yet to see what is and isn't possible.
A first step would be to build a scalable computer and execute simple calculations on it, just like how classical computers started.
According to science you will be ask the computer any question and it will produce the answer instantly and be correct every time. Unless you ask it a paradox than an infinite number of quantum computing cells will implode potentially creating a black hole greater than the mass of the universe.
How do I get into quantum computing? I've been working through the quantum computing for the determined. But there doesn't seem to be a specific field for this. Is it the computer dept? The physics dept? Or something that you have to track down the one guy who actually does it and beg to be his personal assistant?
>>7657985
the quantum science institute at my university focuses mostly on quantum crytography
http://www.iqst.ca/
> Theoretically quantum computers would be able to utilize non-deterministic algorithms to allow infinitely-parallel processing in order to solve an algorithm.
This makes no sense. I suggest reading https://en.wikipedia.org/wiki/BQP.
>>7657963
So it just recreates your moms mouth?
> How do I get into quantum computing? I've been working through the quantum computing for the determined. But there doesn't seem to be a specific field for this. Is it the computer dept? The physics dept? Or something that you have to track down the one guy who actually does it and beg to be his personal assistant?
Depends. Do you want to do theory? If so, what part of theory? There's quantum complexity theory, quantum information theory, theory of quantum programming languages, just to name a few. Do you want to do experimental work? If so, what part?
I suggest reading a couple of research papers that you find interesting and then tracking down the people in those papers to see where they are.
> But the entire field of computer science is built upon deterministic computation and the realm of non-deterministic computation remains relatively unexplored except by a few mathematicians.
This is completely wrong. It is a well-known fact that non-deterministic computing machines are no more powerful than deterministic ones (in the sense that the class of problems that can be solved by both is the same). Reference: https://en.wikipedia.org/wiki/Non-deterministic_Turing_machine.
>>7657780
Quantum computers will mostly be used for specialized tasks - while some algorithms run more efficiently on them than any classical computer could possibly perform, not all problems can be solved by such algorithms, and it's actually expected that some problems will be less efficient to solve on quantum computers.
The biggest application everyone's expecting - going beyond stuff like fast prime factorization, database search, etc - is in quantum simulation. Quantum systems with any significant number of particles are extraordinarily difficult to simulate on a classical computer, with the necessary resources exploding as complexity goes up. A quantum computer, however, is naturally suited to efficiently simulating quantum physics.
This would have substantial - potentially massive, if the simulations are efficient and practical enough - implications for fields like materials science, chemistry, biochemistry, nuclear physics, and anywhere where understanding the precise behavior of many-particle systems like molecules or materials becomes critically important.
Quantum computing would probably have little effect on consumer or commercial hardware; given the difficulties of making qubits stable enough and quantum gates sufficiently low-error, it is unlikely that quantum computers will be cheap enough to be worth how effective they are. Even if this weren't the case, computers would remain largely classical, because for most regular applications ordinary bits are easier to use and more versatile.
>>7657963
That sounds more like a paradox trigger than a quantum computer, but not even a paradox trigger fucks up that badly as long as it's built right.
> The biggest application everyone's expecting - going beyond stuff like fast prime factorization, database search, etc - is in quantum simulation.
We've been expecting this for quite a while now: https://www.cs.berkeley.edu/~christos/classics/Feynman.pdf
>>7658004
I'm looking for information and communications, specifically.
>>7657780
Imagine you have 2 cars. One is extremely fast but can only go to one destination before it must return home, the other is slow but you can drive it around freely.
If you just wanna go to the grocery store which car would you take? The fast one right, you only need to go one place anyways. But what if you need to go to several stores? With the fast car you would have to go to one store, then home, then the next, then home, and so on. If you need to go allot of places its probably better to take the slower car.
Basically classical computers are like the fast car. There good at what they do but also limited in some ways. Quantum computers are, at least on paper, like the slower car. They're not as fast really but they can get certain things done faster because theoretically they can do it in fewer steps.
So just like you wouldn't take the slow car if your just going to one place you'll probably never use a quantum computer for mundane things but it could be used to do some really complex stuff better than a classical computer.
Further question- would quantum computers would probably be coupled with the processors we have nowadays, as a sort of hybrid?
>>7658177
Suppose you had to drive through a maze and lots of traffic and lane closures to boot. Which car is faster then?
>>7658201
It's an analogy. Don't look too deeply into it.
could you integrate quantum parts into an otherwise classical computer, or a mostly classical part of a computer?
If you could, is there a reason you may want to, for regular commercial stuff?
>>7658190
Yes; many quantum algorithms require some processing of classical data, and for general-purpose computing a classical computer is often more suitable. (Also, it's much easier to control a quantum circuit with classical hardware than it is to try and build a fully-quantum general purpose computer.)
>>7657787
Quantum computers use quantum phenomena. For example, you can take a very small, multi-faceted crystal and surround it with a collection of photo-receptive structures. By spinning the crystal, you change the facet that shows to a particular photo-receptor. Each facet will have unique defects, so if you have six sides, you have six distinct bits.
The number of times you rotate the crystal per arbitrary unit of time is your hertz. We can rotate things very fast, and this multiplies your bits by the rotations.
The photons involved display quantum behavior, such as indeterminacy. That is, you can never quite be sure if you've received the photon or not until you make a measurement - at which point, the bit you're looking at appears to be 1 or 0, i.e on or off.
Essentially, every possible register exists at the same time. For this reason, you can't build it like a normal on-off computer. There is no on or off - you turn registers on or off randomly as you observe.
Basically, it takes a human to divine - interpret and give meaning - to the meaningless collection of random 1's an 0's. It reveals the fact that computation occurs naturally. Order is naturally emergent, even as it's like a ghost that doesn't exist.
This is why you hear people talking about the universe being created by observation. The quantum mystics misinterpreted the statement that matter's organization is subjective as saying 'matter is created by observation.' In reality, matter is as real as ever - it's just that in the featureless quantum sea, you can see anything you want.
A quantum computer is more like a set of runes than a digital computer.
Currently there are only a few algorithms that work for a true quantum computer, Shors and a few others. We have a bit to go.
>>7658219
Thanks. I don't know very much about these and it's nice to learn more about them.
>>7658215
Poyentially, yes. But it has to be near 0 degrees kelvin to opperate properly, currently.
Wouldn't it be more likely that future computers will have both quantum and classical components?
>>7658201
Its just a metaphor but I suppose you'd want the slower car for any complex task.
Basically the issue with the fast car, and with classical computers, is that while the are fast they take allot of steps to do things and at some point that starts to make things highly impractical/impossible. Basically once your trip reaches a certain number of stops it doesn't matter how fast your car is, having to go back home in between each stop eventually just becomes too much.
The solution at that point isn't a faster car, its using a car that doesn't have to make those return trips between stops. That's the idea behind a quantum computer, it does things in fewer steps but each step is not necessarily faster than what a classical computer can do.
>>7657996
>>7658019
Sorry, but both of those articles support my claims. Maybe you should read them next time. Granted, if you believed that I was saying that ALL nondeterminsitic algorithms could be expressed via existing theoretical quantum computing models that would be wrong, but I said nothing of the sort.
And second, whether one problem-set is technically a super or sub set of another is completely irrelevant when (a) it would require infinite-length algorithms to map (b) it's never been proven, and (c) computers are engineering things whose capabilities are entirely determined by matters of efficiency, not mathematical extremes.
>is op a faggot?
>YES/NO/MAYBE
Each has 1/3 chance to be right according to quantum computers yet we all know that is not the case.
> Sorry, but both of those articles support my claims. Maybe you should read them next time.
Your claims make no sense.
- You wrote "quantum computers would be able to utilize non-deterministic algorithms". A computer cannot utilize an algorithm. Maybe you meant that it can carry out an algorithm?
- Then you wrote "allow infinitely-parallel processing". You cannot do infinitely many operations on a quantum computer in a finite amount of time. A quantum computer with n qubits can perform one operation on all 2^n possible states in parallel, but you can only read or output one of the states.
- Lastly you wrote "in order to solve an algorithm". Algorithms are not things you solve. They are things you carry out.
>>7658637
Nigger, I'm not even part of this discussion but semantics much?
>>7658637
>when you're not intelligent to actually contribute but want to post anyway
Quantum Annealing could prove to be game changing for AI/Machine Learning. Who knows though.
>>7658223
>a bit to go
No pun intended
