Qubits entanglement and superposition times

[Luveh Keraph]

I am a noob when it comes to QCs, so bear with me.

My understanding (which may be wrong - let me please know if such is the case) is that in order to obtain useful results from a QC the qubits in that device must be kept entangled and in superposition for a finite amount of time, which may vary from a few milliseconds to minutes, or longer, depending on the problem being addressed.

What is the best that has been so far achieved in this respect, both in time and number of qubits? How does it compare with what would be necessary to tackle RSA-2K?

[Phillip Hallam-Baker]

That is a question about quantum computers which is a very different topic.

There are many papers on how many qbits and for how long to tackle RSA at various key sizes. And all of it is pretty much irrelevant as far as PQC goes because the question we have to answer is whether there is any significant chance that a cryptographically relevant QC destroys the global financial system in the next X years. That is an eschaton level threat.

It is over 30 years since I did experimental high energy physics. Looking at the superconducting based QCs, it is pretty clear that they are physics experiments and will not be threatening an eschaton level catastrophe in the next ten years and highly unlikely that they ever will. The basic objection is they do not scale because doubling the power of the machine essentially requires spending twice as much on the refrigerator. Its like spinning plates, the more plates you try to spin, the harder it gets.

There is also the question of whether supeconductivity is a quantum effect or a macro level effect that merely mimics quantum entanglement for a limited number of QBits. Best way to find out is to build them. As I said, experimental physics, proving the theory boys and girls wrong is what we lived for.

Looking at trapped ion based machines, they are not yet a practical basis for computation. But if they ever become practical, they can be made with conventional VLSI and do not rely on having a big fridge. And while there are obstacles to overcome before it is known such a device is even possible, there is certainly a possibility, 1% say, someone builds one within the next decade. And if they do, it will go from ten to millions of QBits very quickly because we already solved the VLSI scaling problem.

So as far as the eschaton level threat goes, there is a small but real threat a CRQC destroys civilization unless we lock the cryptography down. So lets do that. 

That said, there is a very real possibility that the theory boys and girls have it completely wrong and there are limits to entanglement. So far we have three photons entangled. But three is not thousands or millions.

This is science and science is about methodological doubt, not certainty.