Quantum Computing

[eloic]

First proof of quantum computer advantage
Quantum computers promise to revolutionize the future of computing. Scientists have now demonstrated for the first time that quantum computers do indeed offer advantages over conventional computers. They developed a quantum circuit that can solve a problem that is unsolvable using any equivalent classical circuit.

https://www.sciencedaily.com/releases/2018/10/181018141107.htm

 

[Deleted member 2197]

Quantum Computing Is Mathematically Better, But Not Always

It's always been assumed that quantum computing is better — at least to the layperson. But IBM research scientists have now actually proven mathematically that quantum computing is faster than a classical computer for certain problems.

Sutor said this proof is an extremely important milestone, as it will be a foundation for building the rest of the formal structure of quantum computers — how they are coded, built, and what choices are made around algorithms and how they are applied. But it will also provide guidance on when quantum computing might be the option, or whether classical computing will still be sufficient, he said.

To understand the significance of the proof, it's important to understand the basic computational unit in quantum computing is a quantum bit — a qubit, which unlike a classical bit that is always 0 or 1, can take on many other additional values. The potential computational power doubles every time a qubit is added through entanglement, and the qubits together with the operations applied to them are called a circuit.

Qubits aren't perfect in that they have small error rates and only exist for a certain length of time before they become chaotic. This is known as the coherence time, and it means there only so many operations that can be done before the time limit is reached. The number of operations performed is depth, and the overall depth of a quantum circuit is the minimum of all the depths per qubit. The math proves that certain problems need only a fixed circuit depth when done on a quantum computer, no matter how the number of inputs increases, whereas a classical computer requires the circuit depth to grow larger as inputs increase for the same problem.

This limited depth means IBM scientists are most interested in what can be done with short-depth circuits because they are practical for implementing quantum algorithms and demonstrating quantum computing has an advantage over a classical approach. The mathematical proof shows that fault tolerant quantum computers will do some things better than classical computers can — not all. “It's a little subtle I think sometimes,” said Sutor. “But that's a very important distinction.”

https://www.eetimes.com/document.asp?doc_id=1333900&_mc=RSS_EET_EDT

 

[Orion]

I think people like Wofram believe it will likely prove physically impossible to significantly scale quantum computers.

As for optical it likely has too large components. Nanomagnet computers approach theoretical energy efficiency physical limit and could have small components.

Optical could still serve for interconnects in other types of computer.

 

[eloic]

Google may have taken a step towards quantum computing 'supremacy' (updated)
Or maybe not -- the paper has been unpublished and its results are disputed.

link

 

[Davros]

Electrical quantum computing is propagandised to keep people's eyes off the prize.

Just to be clear, your not suggesting there is a conspiracy to stop people developing optical computers are you ?

 

[MfA]

Just to be clear, your not suggesting there is a conspiracy to stop people developing optical computers are you ?

No, I suspect there are concerted efforts to selectively fund and promote electrical quantum computing to delay research into optical quantum computing. You can't stop it, but you could slow it down.

PS. here's the Google paper BTW. Not too interesting, some quantum logic running logic with no practical application which can't be simulated at decent speeds ... but not a proven NP problem AFAICS. So what's the point in proving it's faster than simulation? There might be polynomial algorithms out there to produce the same outputs.

 

[Davros]

Well optical is harder It makes sense to get electrical working first.
I dont think anyones funding electrical to slow down optical - you sort of are suggesting some sort of conspiracy

 

[eloic]

Agaaaaaaaaaain:
https://www.space.com/quantum-computer-milestone-supremacy.html

'Supremacy' Achieved: Quantum Computer Notches Epic Milestone
By Mike Wall 11 hours ago Tech

Quantum supremacy has apparently arrived.

Google’s quantum supreme cryostat, with the Sycamore quantum computer inside.
(Image: © Erik Lucero/Google)
We have just entered the age of quantum supremacy, a new study suggests.

For the first time, a quantum computer has solved a problem that a traditional computer, for all practical purposes, cannot, researchers reported today (Oct. 23) in the journal Nature.

 

[Davros]

It's another headline thats full of a lot of hyperbole
ps: if the computation would take 10,000 years on a classical supercomputer how do we know the quantum computer got the right answer ???

 

[Karamazov]

because the quantum computer said so, and he is more intelligent than us all

 

[pcchen]

It's another headline thats full of a lot of hyperbole
ps: if the computation would take 10,000 years on a classical supercomputer how do we know the quantum computer got the right answer ???

IBM argues that Google didn't properly optimize the computation on classical computers: https://www.ibm.com/blogs/research/2019/10/on-quantum-supremacy/
The claimed "computation would take 10,000 years" was more like 2.5 days. Of course, a few minutes is still much faster than that.

To answer your question on how to know if the quantum computer got the right answer, there are some problems which are hard to solve, but relatively easy to verify.
For example, if I give you an AES encrypted message and the original message, such as, "Hello World" and "AMXKSCJDOIUDFA" (not real encrypted message), it'd be very difficult to find the key, which is the point of an encryption algorithm. However, if someone gives you key, it's very easy to verify the encrypted message is indeed encrypted from the original message with the key.

 

[Deleted member 2197]

A carefully chosen problem to make Quantum computers look good?

John Preskill, a CalTech physicist who originated the phrase 'quantum supremacy' in a 2012 paper, recently wrote about Google's experiment in Quanta Magazine (the experiment results were leaked early). Again, hearty congratulations were offered, but Preskill questioned the problem set and comparison with classical computer capabilities.

Specifically, Preskill noted that the problem chosen by Google "was carefully chosen just for the purpose of demonstrating the quantum computer's superiority," and the answer provided at the end of the calculation wasn't very worthwhile. The biggest hope from this is that now this milestone is behind them, the research team at Google can search for more useful applications in which Quantum Computers can shine.

https://hexus.net/tech/news/cpu/136028-ibm-disputes-googles-quantum-supremacy-claim/

 

[pcchen]

A carefully chosen problem to make Quantum computers look good?

https://hexus.net/tech/news/cpu/136028-ibm-disputes-googles-quantum-supremacy-claim/

Of course, a real quantum computer is only good at certain problems, so I think it's ok. And you have to start somewhere
The important point of Google's paper is that they actually made something at calculating something which is much faster than anything else. This "proves" that it's possible that a quantum computer is better at calculating things.
This is important because before this all we have is something like "IBM made a quantum computer able to factor 15 into 3x5" which is of course trivial for any classical computer (or human).

On the other hand, this is not to say that RSA (or other similar public key cryptography) is now in danger. It still takes a quantum computer with maybe millions of qubits to solve 2,048 bits RSA, and now we only have an extremely noisy 53-qubit one. And people are already researching into the so-called quantum resistant cryto algorithms.

 

[Davros]

A carefully chosen problem to make Quantum computers look good?

Please dont use black text you cant see it with the dark theme

 

[Deleted member 2197]

Please dont use black text you cant see it with the dark theme

Changed to "none" font color and bolded. Should be visible ...

 

[Davros]

thanks

 

[eloic]

 

[eloic]

Ok, mfs, Google's Willow is here.

 

[pcchen]

One thing to note is that NIST already recommended people to prepare for migrating to post-quantum encryptions. I think people really should consider the possibility that such migration might be necessary in as few as 5 years.

 

[IQandHDR]

Last time I read, it would take like 20.000.000 Qubits being in superposition for 8 hours to break current level of encryptions?

Seems like a target far away still?