Well it turns out that to launch Mail you have to configure it.
So what's the point of using another mail client for the regular Joe, if he has already configured Mail?
I'm considering running linux on my mac lately...
I disagree. Presumably, once something like this is working, the performance can and will be improved upon. But you will need to approach unit success probability for ever larger computations, to keep the overall success probability close to 1. In that sense I feel this is not scalable.
Also, a classical computer can only fully simulate something like 12 qubits (maybe even less, like 8, I don't remember), so it doesn't have to scale that much to be useful for quantum simulations. For decrypting RSA, then yes, scaling to 1024+ bits would be a bit more challenging to say the least. I agree on the limitations of classical computing, and I agree that photon QC may be useful for solving small enough problems, but I think that is not the issue. My point is that with this scheme you cannot improve the complexity class of any problem.
Say you have at any given time n photons flying around and you keep throwing photons in as you measure them, so the total amount of photons involved in the final computation is N (maybe much larger than n). Still, the relevant number in order to compare with a classical computer (CC) will be n, since that is what the CC needs to remember at any given time, and that is what determines the difficulty to simulate time evolution. So scaling N doesn't make a difference for a classical computer. That can be afforded with a linear increase in time. What a CC cannot deal with efficiently is scaling n, and there is where QC should improve. As I said, to achieve that with photons you need to increase the success probability of single photon emission to arbitrarily close to 1.
That sounds impressive, but still, I don't see how can that be scalable. Performing a quantum computation with n qubits you're gonna need to generate n photons simultaneously (apart from error correcting codes, which I'm gonna neglect), so the probability of success will go as
P=p^n -> 0
where p is the probability of success in generating a single one. Therefore, as n grows large, the success probability goes to zero. In other words, you'll need better success probability for ever larger computations. In that sense I think that this is not scalable.
It is not my intention to underestimate the relevance of your achievement, which may be useful for many other applications such as cryptograpghy... but the way I see it, not QC.
Not really...
If at any time during the computation you have a limited amount of photons and as you measure them you keep throwing more in... well... that's not gonna work, because the amount of information you would need to describe such system at any given time would not increase with the number of qubits that the computation takes, therefore it can be simulated in a classical computer therefore is not giving you anything more than classical power.
The power of quantum computers will only be harnessed when all qubits coexit at a given time. That you cannot simulate efficiently as the number of qubits grow, and that's giving you the power of QC. This is why I am reluctant that scalable quantum computation will occur with photons... It's too difficult to create single photons on-demand.
Spin is a quantum property of electrons, so in some sense, it can be used to store quantum information. However, spintronics is (for the time being) looking at another direction. For quantum information you want to look at the state of a single spin, not a whole bunch of them rolling down a wire, 'cause then you loose the essential quantum-cool-features.
Slashdot Mirror
Well it turns out that to launch Mail you have to configure it. So what's the point of using another mail client for the regular Joe, if he has already configured Mail? I'm considering running linux on my mac lately...
That sounds impressive, but still, I don't see how can that be scalable. Performing a quantum computation with n qubits you're gonna need to generate n photons simultaneously (apart from error correcting codes, which I'm gonna neglect), so the probability of success will go as P=p^n -> 0 where p is the probability of success in generating a single one. Therefore, as n grows large, the success probability goes to zero. In other words, you'll need better success probability for ever larger computations. In that sense I think that this is not scalable. It is not my intention to underestimate the relevance of your achievement, which may be useful for many other applications such as cryptograpghy... but the way I see it, not QC.
Not really... If at any time during the computation you have a limited amount of photons and as you measure them you keep throwing more in... well... that's not gonna work, because the amount of information you would need to describe such system at any given time would not increase with the number of qubits that the computation takes, therefore it can be simulated in a classical computer therefore is not giving you anything more than classical power. The power of quantum computers will only be harnessed when all qubits coexit at a given time. That you cannot simulate efficiently as the number of qubits grow, and that's giving you the power of QC. This is why I am reluctant that scalable quantum computation will occur with photons... It's too difficult to create single photons on-demand.
Spin is a quantum property of electrons, so in some sense, it can be used to store quantum information. However, spintronics is (for the time being) looking at another direction. For quantum information you want to look at the state of a single spin, not a whole bunch of them rolling down a wire, 'cause then you loose the essential quantum-cool-features.
Aerodynamics has no use in open space