Re:IBM casts doubt on Google's claims of quantum supremacy
by sanxiyn on Wednesday October 23, 2019 @09:16AM
(#21334283)
The key claim is linear scaling. IBM: "The runtime of our simulation method scales approximately linearly with the circuit depth". That directly contradicts what Google wrote in the paper: "... algorithm becomes exponentially more computationally expensive with increasing circuit depth".
Re:IBM casts doubt on Google's claims of quantum supremacy
by altares on Wednesday October 23, 2019 @09:29AM
(#21334398)
Supremacy does not mean simply faster. That is already proven. It means impossible to do on a classical computer.
Re:IBM casts doubt on Google's claims of quantum supremacy
by djhaskin987 on Wednesday October 23, 2019 @09:39AM
(#21334482)
I'm sure you mean infeasible. "Impossible" goes back to a problem's decidability. If a problem is undecidable, there's no way you're going to find its solution by switching computational models; classical computers can do anything a quantum computer can do, just "slower".
Re:IBM casts doubt on Google's claims of quantum supremacy
by akvadrako on Wednesday October 23, 2019 @10:45AM
(#21335135)
Not if they use up all the energy in the visible universe or they need so much power they collapse and form a black hole.
Real quantum computers should be able to do calculations that are that powerful, assuming we can prove the best classical algorithms are really exponential.
Re:IBM casts doubt on Google's claims of quantum supremacy
by orbifold on Wednesday October 23, 2019 @09:36AM
(#21334451)
The quantum circuit based simulation actually has much worse accuracy. A classical computer can simulate the class of circuits they are proposing with linear increasing runtime (as opposed to exponential as Google claims) and _much_ higher accuracy according to IBM.
Re:IBM casts doubt on Google's claims of quantum supremacy
by ascii_only on Wednesday October 23, 2019 @09:31AM
(#21334408)
It is difference between with this quantum chip we can brake encryption that would take 100000 years in 30 minutes or it would take us 50 years.
I don't see the argument here