The end of quantum computing?

https://phys.org/news/2022-12-uncovers-limitations-entanglement.html

This piece seems to suggest it is near impossible to coordinate all those qubits for something even as small as a 100 qubit machine.
Problem is we need more like a million qubits to start getting answers to the really big questions so could this be saying there will be no huge number quantum bit computer?

@sonhouse

I’d love to read the linked page, but the server won’t let me opt out of ‘personalized’ advertising cookies. Can you please summarize?

From the source, HTML tags and all:

<p>Researchers at Tsinghua have recently carried out a study exploring the possible reasons why the reliable and efficient detection of <a href="https://phys.org/tags/entanglement/" rel="tag" class="textTag">entanglement</a> in complex and "noisy" systems has often proved to be very challenging. Their findings, published in <i>Physical Review Letters</i>, hint at the existence of a trade-off between the effectiveness and efficiency of entanglement detection methods.</p>
<p>"Over 20 years ago, <a href="https://journals.aps.org/pra/abstract/10.1103/PhysRevA.58.883">researchers discovered</a> that most quantum states are entangled," Xiongfeng Ma, one of the researchers who carried out the study, told Phys.org. </p>
<p>"This means that, for example, if we managed to construct a 100-qubit system, say, a superconducting or ion-trap quantum computing system, this system will evolve for a while, during which the qubits extensively interact with each other. Of course, there will be errors, so to maintain a good coherent control, we reasonably isolate the system from the environment. As long as the purity (quantifying the effectiveness of our isolation effort) is not exponentially small with the number of qubits, the system is highly likely to be entangled."</p>
<p>While entanglement might theoretically seem fairly simple to realize, achieving it in experimental settings is in fact very difficult. Studies showed that it is particularly difficult in large quantum systems, such as <a href="https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.120.260502">systems comprised of 18-qubits</a>. The key objective of the recent work by Ma and his colleagues was to gain a better understanding of the challenges associated with the detection of entanglement in large systems.</p>
<p>"Researchers gradually realized that while the preparation of entangled state for a large system might be easy, the entanglement detection could be very challenging in practice," Ma explained. "In our work, we establish a mathematical formulation to quantify the effectiveness of an entanglement detection method. We employ a proper quantum state distribution, use the ratio of detectable entangled state to quantify its effectiveness, and also quantify the efficiency of an entanglement detection method by the number of observables needed for this method."</p>

<p>Ma and his colleagues first examined what is arguably the most straightforward entanglement detection <a href="https://phys.org/tags/protocol/" rel="tag" class="textTag">protocol</a> available today, known as entanglement witnesses. They showed that this protocol's ability to detect entanglement decreases by a double exponential value as the system gets larger.</p>
<p>The researchers later found that this reduction in effectiveness linked to a system's size also affected other entanglement detection protocols. After a series of theoretical considerations, they were able to extend their observations of the entanglement witnesses method's performance to arbitrary entanglement protocols that rely on single-copy quantum state measurements.</p>
<p>"For a random state coupled with the environment, any entanglement detection protocol with single-copy realization is either inefficient or ineffective," Ma said. "Inefficient means the protocol relies on measuring an exponential number of observables and ineffective means the success rate of entanglement is double exponentially low."</p>
<p>Essentially, Ma and his colleagues showed that to observe entanglement on a large-scale, researchers must be able to control all interactions in a system with high precision and know almost all information about them. When there is a lot of uncertainty about the system, therefore, the probability of detecting its entanglement is very small, even if one is almost certain of its occurrence.</p>
<p>"We proved that no entanglement detection protocols are both efficient and effective," Ma explained. "This may help the design of entanglement detection protocols in the future. Meanwhile, detecting large-scale entanglement could be a good indicator for comparing different quantum computing systems. For example, when a lab team claim they build a hundreds-of-qubit system, they should detect entanglement. Otherwise, they have not controlled the system well enough."</p>
<p>Overall, the findings gathered by this team of researchers highlight the existence of a trade-off in the efficiency and effectiveness of existing entanglement detection methods. In addition, they offer valuable insight about the reasons why detecting entanglement in large-scale and noisy quantum systems is so difficult.</p>
<p>"Our result does not prevent us from designing a protocol that is both efficient and effective when the system is well-controlled (i.e., the coupled environment is relatively small)," Ma added. "Currently, we only have entanglement detection protocols that work well for pure states, such as entanglement witnesses, and protocols that work for large environments at the expense of exponential cost. We noticed that an entanglement detection protocol that works for moderate environment size with relatively low cost is still missing, and we would now like to try to develop one."

@sonhouse

Maybe new "protocols" can be devised.

The article mentions error control thus:

Of course, there will be errors, so to maintain a good coherent control, we reasonably isolate the system from the environment.

This is not the only way to control error, of course. The other way is redundancy, which is built into things like QR codes, I believe, so that the code on a package can be correctly scanned even if a portion of the sticker is defaced or covered up.

I doubt this is the end of quantum computing.

@soothfast said
I doubt this is the end of quantum computing.

It's phyzzog clickbait. The only thing it'll be the end of is honest science reporting.

@shallow-blue said
It's phyzzog clickbait. The only thing it'll be the end of is honest science reporting.

What is dishonest about it, exactly?

I'd agree if the title of the article were like the title of this thread, but it is not. "Study uncovers existing limitations in the detection of entanglement" seems fairly restrained and accurate.

Incidentally this thread caught my eye because I was already looking at getting some rigorous understanding of quantum information theory. A highly-rated, modern book that is relatively accessible is here:

https://markwilde.com/qit-notes.pdf

Mark Wilde is the author, and while the 2nd edition of the book is published by Cambridge University Press, he has been given permission to have the entire text on his own website.

The book restricts itself to finite-sized systems, which obviates the need to brush up on Hilbert spaces and deal in complicated areas of functional analysis. A reader just needs to have a good grasp of probability theory (discrete distributions from the looks of it) and linear algebra.

EDIT: The pdf linked to above is a draft of the published book, which is titled "Quantum Information Theory."

At https://en.wikipedia.org/wiki/Quantum_information I find this to be a clear characterization of quantum information theory as it relates to quantum physics (or quantum mechanics) as a whole:

Quantum mechanics is the study of how microscopic physical systems change dynamically in nature. In the field of quantum information theory, the quantum systems studied are abstracted away from any real world counterpart. A qubit might for instance physically be a photon in a linear optical quantum computer, an ion in a trapped ion quantum computer, or it might be a large collection of atoms as in a superconducting quantum computer. Regardless of the physical implementation, the limits and features of qubits implied by quantum information theory hold as all these systems are mathematically described by the same apparatus of density matrices over the complex numbers. Another important difference with quantum mechanics is that, while quantum mechanics often studies infinite-dimensional systems such as a harmonic oscillator, quantum information theory concerns both with continuous-variable systems and finite-dimensional systems.

A definition of quantum Shannon theory: "Quantum Shannon theory describes communication in terms of abstract quantum systems, representing the degrees of freedom used to carry information."

@moonbus said
@sonhouse

I’d love to read the linked page, but the server won’t let me opt out of ‘personalized’ advertising cookies. Can you please summarize?

Clearing your browser removes all cookies no?

I wouldn't mind any explanation about cookies because I'm still confused about it.

@soothfast said
From the source, HTML tags and all:

<p>Researchers at Tsinghua have recently carried out a study exploring the possible reasons why the reliable and efficient detection of <a href="https://phys.org/tags/entanglement/" rel="tag" class="textTag">entanglement</a> in complex and "noisy" systems has often proved to be very challenging. Their findings, published in <i>Physical ...[text shortened]... onment size with relatively low cost is still missing, and we would now like to try to develop one."

Thanks for the full text.

Researchers at MIT have developed a programming language, TWIST, which apparently reliably detects which qubits are entangled and which are 'pure.'

https://spectrum.ieee.org/quantum-programming-language-twist

I am optimistic that this issue can be resolved and the full power of quantum computing soon released.

@booger said
Clearing your browser removes all cookies no?

I wouldn't mind any explanation about cookies because I'm still confused about it.

Each browser (Safari, Chrome, Firefox, etc.) has the facility to delete cookies, but you have to know the exact sequence of clicks to turn the function on to automatically clear cookies after a certain period of time or at every restart.

Cookies are data deposited on your computer by web sites you visit. They store information which allow web servers to send you targeted information (such as adverts), but also tailor the web site content to your browser's settings (which character sets and fonts you prefer, and such like). Cookies may also record the date and time of your previous visit(s), how long you tarried, what search words you key-in, and basically anything you did on that web site could potentially be stored, for later retrieval. The purpose of this is to feed you more and more of the same; whatever you watched last time, you get more of next time. If you search for news items on right-wing conspiracies, you'll automatically get force-fed more and more news items on right-wing conspiracies. If you search on John Lennon's dirty socks, you'll get more and more adverts from people selling John Lennon's dirty laundry.

@moonbus
The problem is we won't get very far with just a dozen or two Qubits, we need thousands and the gist of that article says it is more like trying to hold strands of cooked spaghetti end to end to transmit light through for instance.
The more Qubits there are, according to this piece, the more difficult it gets to ensure proper entanglement.
I wonder what they will do for a work around.
One way I see is more robust Qubits, maybe photons, I see they are working with entangle photons and maybe they will be more inherently stable.

Just a guess though.

@sonhouse said
@moonbus
The problem is we won't get very far with just a dozen or two Qubits, we need thousands and the gist of that article says it is more like trying to hold strands of cooked spaghetti end to end to transmit light through for instance.
The more Qubits there are, according to this piece, the more difficult it gets to ensure proper entanglement.
I wonder what they will ...[text shortened]... working with entangle photons and maybe they will be more inherently stable.

Just a guess though.

Consider traditional computer processors. Generally, the more transistors that can be crammed onto a microchip, the greater the processing power. Since there's a technological limit to how small a transistor can be, to increase processing speed beyond what a single chip can provide it will be necessary to put two or more chips into a single computing system and employ parallel processing algorithms. Right?

So if we can't make a single quantum processor that has more than a couple dozen qubits, can't multiple quantum processors be combined in parallel to one another to make a single quantum computer with the desired power?

@Soothfast
I think that was the gist of the piece, the more qubits the less stable they are as an group. I don't think it is like say putting in a USB hub and now you have 4 ports instead of 2 or some such. Those qubits interact and I think that is the implication, the noise in the system goes up so much qubit integrity cannot be maintained.
Of course, scientists and engineers are sneaky and maybe some new development will trump this piece.
Been done before🙂

@moonbus

Thanks for the explanation.