Concerning Quantum computers and information

[naggy]

I only recently 'discovered' this quantum computer business and quantum information.

As I see it, this seems like a drastic departure from regular 'classical' computers. So if we ever get to the point of using these computers, won´t they have to rediscover the entire field of computer-science?

The memory works completely different, so all algorithms must change? Just look at this stuff, it's crazy, called Shors algorithm, it only factors an integer into it's prime components.

http://en.wikipedia.org/wiki/Shor's_algorithm

I've only learned introduction to Quantum Mechanics, but still I don´t understand half of this stuff.

 

[f95toli]

As I see it, this seems like a drastic departure from regular 'classical' computers. So if we ever get to the point of using these computers, won´t they have to rediscover the entire field of computer-science?

People have been working on algorithms etc for quantum computers for quite a while, so a lot of work has already been done.
Also, much of underpinning theories for quantum algorithms goes back to something that is known as reversible computing, and that field has been around for 40 years or so.

The "theoretical" side of quantum computing is much, much more advanced than the practical.

 

[naggy]

People have been working on algorithms etc for quantum computers for quite a while, so a lot of work has already been done.
Also, much of underpinning theories for quantum algorithms goes back to something that is known as reversible computing, and that field has been around for 40 years or so.

The "theoretical" side of quantum computing is much, much more advanced than the practical.

So will a computer-scientist today have to relearn a lot when we'll start using these computers?

 

[Tao-Fu]

I think it is more likely that quantum processing units will be used as specialized processors to carry out special tasks that benefit from ultra-parallel computation. Quantum computers are probabilistic not deterministic, so you can't give a set of well-defined inputs and obtain a unique output in general. This does pose a problem for traditional programs as we know them that take the contents of registers and modify them in well-defined and predictable ways. To get useful output from a quantum computer you need to measure its state and get classical information out of it (a probabilistic and irreversible process). A classical computer can efficiently verify the output from the quantum processing unit for Grover's search algorithm and Shor's prime factorization algorithm. In fact I am not sure if a quantum computer can even verify its own output. Usually, it is understood that the classical output of the quantum computer will be used in a task in a classical computer (this includes verification of the output and its use in a classical algorithm).

 

[ZapperZ]

I think it is more likely that quantum processing units will be used as specialized processors to carry out special tasks that benefit from ultra-parallel computation. Quantum computers are probabilistic not deterministic, so you can't give a set of well-defined inputs and obtain a unique output in general. This does pose a problem for traditional programs as we know them that take the contents of registers and modify them in well-defined and predictable ways. To get useful output from a quantum computer you need to measure its state and get classical information out of it (a probabilistic and irreversible process). A classical computer can efficiently verify the output from the quantum processing unit for Grover's search algorithm and Shor's prime factorization algorithm. In fact I am not sure if a quantum computer can even verify its own output. Usually, it is understood that the classical output of the quantum computer will be used in a task in a classical computer (this includes verification of the output and its use in a classical algorithm).

Er... did you read this report before you wrote this?

http://www.wired.com/wiredscience/2010/01/quantum-computer-hydrogen-simulation/

Reference: B.P. Lanyon et al, Nature Chemistry v.2, p.106 (2009).

Zz.

 

[Tao-Fu]

No. But that's about quantum simulation, not quantum computing, exactly. There's no problem if you wish to use a quantum system to simulate another quantum system. This isn't quite what people usually mean by a quantum computer, however. They want something that initializes a quantum register of qubits and provides an output into an equivalently sized classical register of bits.

 

[ThomasEdison]

From the link supplied by PF mentor:
"These uses might include the ability, without trial and error, to design complex chemical systems and advanced materials with properties never before seen."

That sounds promising.

 

[ZapperZ]

No. But that's about quantum simulation, not quantum computing, exactly. There's no problem if you wish to use a quantum system to simulate another quantum system. This isn't quite what people usually mean by a quantum computer, however. They want something that initializes a quantum register of qubits and provides an output into an equivalently sized classical register of bits.

Er.. this is a play on words. They ARE computing the hydrogen molecules energy state! That's what they mean when they say "simulate". Last time I checked, ordinary computers do that as well!

I suggest you read the actual paper before making any further remarks on this.

Zz.

 

[Phrak]

Er... did you read this report before you wrote this?

http://www.wired.com/wiredscience/2010/01/quantum-computer-hydrogen-simulation/

Reference: B.P. Lanyon et al, Nature Chemistry v.2, p.106 (2009).

Zz.

Why do you bring-up this article as if it is something everyone should know about after having read the OP's link to Wikipedia? The Wikipedia article has no mention of it.

Is this on the required reading list of anyone commenting on quantum computing? Is Tao-Fu somehow at fault for this oversite?

 

[ZapperZ]

Why do you bring-up this article as if it is something everyone should know about after having read the OP's link to Wikipedia? The Wikipedia article has no mention of it.

Is this on the required reading list of anyone commenting on quantum computing? Is Tao-Fu somehow at fault for this oversite?

1. He made a presumption about the ability of quantum computers.

2. I then asked if he wrote that after he has read this article, which in my opinion, clearly falsifies the assertion.

3. He/she said no, that he/she is not aware of the article, but still made some assertion about it not being a "computer", which I thought was rather silly.

4. I asked him/her to read the paper then to make sure he get his info up-to-date.

5. It is not my problem that the Wikipedia doesn't have it. Is it a total surprise that a Wikipedia article doesn't have accurate or up-to-date information?

Part of learning is the discovery of new information that you didn't know before. So what are you objecting to? I didn't require him/her to be aware of it. I gave information that is pertinent to the discussion.

Zz.

 

[f95toli]

Er.. this is a play on words. They ARE computing the hydrogen molecules energy state! That's what they mean when they say "simulate". Last time I checked, ordinary computers do that as well!

I suggest you read the actual paper before making any further remarks on this.

Zz.

I think I must agree with Tao-Fu here. What the Brisbane group has made is not really what most people would refer to as a quantum computer. It has more in common with the analogue computers (based on op-amps and RCL circuits) people used to solve problems in physics during the 60s and 70s than to what we today mean by a "generic", programmable, computer.
To some extent it IS of course a play on words, and it goes back to the old discussion about whether or not analogue computers are really computers. But there is also a more practical side, as far as I can tell there is no way to e.g. implement error correcting algorithms in this scheme (at least not beyond re-focusing pulses) meaning there are inherit limitations when it comes to the size of the system that can be simulated.

 

[ZapperZ]

I think I must agree with Tao-Fu here. What the Brisbane group has made is not really what most people would refer to as a quantum computer. It has more in common with the analogue computers (based on op-amps and RCL circuits) people used to solve problems in physics during the 60s and 70s than to what we today mean by a "generic", programmable, computer.
To some extent it IS of course a play on words, and it goes back to the old discussion about whether or not analogue computers are really computers. But there is also a more practical side, as far as I can tell there is no way to e.g. implement error correcting algorithms in this scheme (at least not beyond re-focusing pulses) meaning there are inherit limitations when it comes to the size of the system that can be simulated.

But the key issue here is that one can, in fact, come up with a quantum computation algorithm, which is a crucial aspect of any kind of computation[1]. In this particular case, it is an algorithm based on the quantum mechanical aspect of the superposition of light in various path/states. I would seriously doubt that this isn't as aspect of quantum computation.

And the fact that they could get such a high accuracy in the result using such algorithm should address directly the issue that was brought up here.

Zz.

[1] A.M. Childs and W. van Dam, Rev. Mod. Phys. v.82, p.1 (2010).