Trying to get my head round quantum computers

kwerk · 27 Mar 2018 at 19:09

QlimaxUK said:
The question everyone want to know.. when will these be ready for the domestic market?

now I know these things are big, complex and require crazy low cooling to work at all so will be a long long way off so are we talking 20 30 40+ years.. or never? make your predictions now!

When we have access to exotic matter. I think we'll just send jobs to a remote cloud like those old timey mainframes instead of home ones.

Mr C · 27 Mar 2018 at 19:24

I remember reading a large article about it in The Economist's print edition. It seemed quite comprehensive at the time. I think from this i got the gist of what they are trying to do.

I believe this link is the same:

https://www.economist.com/news/essays/21717782-quantum-technology-beginning-come-its-own

rubberduck · 27 Mar 2018 at 19:28

QlimaxUK said:
The question everyone want to know.. when will these be ready for the domestic market?

now I know these things are big, complex and require crazy low cooling to work at all so will be a long long way off so are we talking 20 30 40+ years.. or never? make your predictions now!

I don't think you would actually want one, even if they could be prodced to a point where they could be purchased by the end user. The cost of the cooling would be prohibitively expensive but furthermore they just wouldnt be able to run applications or games - and even if they could they would be slower than a conventional computer. They are only useful for solving specific problems where conventional number crunching is ridiculously inefficient to the point of almost requiring infinite computational power. Outside of these specific problems they would fail badly. I doubt if they could even run Crysis maxed.

Fusion · 27 Mar 2018 at 20:20

rubberduck said:
I doubt if they could even run Crysis maxed.

I think they probably could

v0n · 27 Mar 2018 at 20:32

Not only they can't really ever run Crysis but you have to feed the problem to solve using regular computer and interpret the results on regular computer. It's purely multitude of "is it positive or negative" things done in the same time, expotentially with each additonal cubit, so by the time you pose the problem to something like 200 cubits you pretty much exceed traditional computational power of all personal computers in the world.
In other words what you most likely to do with it is serious number crunching - buyers of quantum computers like Lockheed Martin or other gov contractors like Airbus will say molecule model computations but think cryprography...

Tuppy_Glossop · 27 Mar 2018 at 22:40

You'll need new Linpack libraries for the parallelisation.

rubberduck · 27 Mar 2018 at 22:46

So how do you write code for a quantum computer then?

Mr Badger · 27 Mar 2018 at 23:03

rubberduck said:
So how do you write code for a quantum computer then?

If you do it right, then it programs itself. Plus, if you think you understand quantum mechanics, then you don't understand quantum mechanics.

LeMson · 28 Mar 2018 at 00:41

Coin miners wet dream then lolz

OspreyO · 28 Mar 2018 at 01:38

Someone asked me if knew how quantum computers work...
.....I told him, yes and no....

mrbazmondo · 28 Mar 2018 at 01:50

kwerk said:
Now Imagine person B on the same trail. Except they have a time freezing machine! They can now explore both forks of the trail for a short distance, at the same relative time, find out how those new trails intersect with other trails in the system, then then pick the one with the best probability of optimal future trail entropy (and hence efficiency). Now they are allowed to return to the fork, start time again and decide which fork to take for real. This all happens in the same 1 second period person A took. That's quantum computing. The exploration of both trails at the same time is a qbit. Technically the hiker is on both of them until he meets a new fork and has to decide which one he's actually on so he can assess the next fork. It's all wave functions and probabilities until you pass it back to the normal computer.

So just like the karnak character on Inhumans then?

SPG · 28 Mar 2018 at 13:34

QlimaxUK said:
The question everyone want to know.. when will these be ready for the domestic market?

now I know these things are big, complex and require crazy low cooling to work at all so will be a long long way off so are we talking 20 30 40+ years.. or never? make your predictions now!

Some Military research complex will have one in a well advanced state i would imagine.

rubberduck · 28 Mar 2018 at 22:50

Hello, thanks for the input. It seems "Grover's Algorithm" is what I was after to better understand quantum computers - plenty of youtube videos out there ranging in quality.

Out of interest. Can you overclock a quantum computer? I would assume yes but wanted to check.

FortuitousFluke · 28 Mar 2018 at 22:56

adolf hamster said:
no that's a different scenario, the common misconception here is that the 3 states are: alive, dead, both

however there is a 4th state that is always true: bloody furious

Is he furious because he just opened a box and found it contained his dead cat?

Rroff · 28 Mar 2018 at 22:57

rubberduck said:
Out of interest. Can you overclock a quantum computer? I would assume yes but wanted to check.

I don't think you understand how it would work. (Though technically there are parts of it that you could).

rubberduck · 28 Mar 2018 at 23:07

Rroff said:
I don't think you understand how it would work.

Well it still needs to execute operations and thus I would assume that there can be variability in speed of operation based on the architecture of the quantum computer. After all, the resolution of the wave function is not the limiting step in speed as I believe this happens essentially instantaneously.... Even at a distance as per Eistein's famous EPR paper which tried to discount such a principle but was later discredited by Bell's inequality through theory initially and more recently multiple experiments to discount the 'hidden information' theory favoured by Einstein et al.

So the question remains, can you overclock a quantum computer, please explain your answer and rationale.

adolf hamster · 28 Mar 2018 at 23:29

FortuitousFluke said:
Is he furious because he just opened a box and found it contained his dead cat?

no the cat is furious because you locked it in a box

mrbazmondo · 29 Mar 2018 at 00:32

You can either understand quantum computers or use them you can't do both.

Rroff · 29 Mar 2018 at 00:44

rubberduck said:
Well it still needs to execute operations and thus I would assume that there can be variability in speed of operation based on the architecture of the quantum computer. After all, the resolution of the wave function is not the limiting step in speed as I believe this happens essentially instantaneously.... Even at a distance as per Eistein's famous EPR paper which tried to discount such a principle but was later discredited by Bell's inequality through theory initially and more recently multiple experiments to discount the 'hidden information' theory favoured by Einstein et al.

So the question remains, can you overclock a quantum computer, please explain your answer and rationale.

There is a lot of confusion there over local and remote "hidden" variables - IIRC Bell and more recent studies have tended to make it less likely there are local hidden variables but in many cases suggestive that remote "hidden" variables exist. Studies recently on entangled pairs also tend to indicate that information in this scenario would have a latency which in terms of the physical universe mean travelling at between 80 and more than 10,000x the speed of light.

You can adjust the rate you are invoking the system but the way quantum mechanics are used themselves don't have a tickrate per operation like a classical computer to produce an outcome from that part.

Trying to get my head round quantum computers

Wise Guy