10
$\begingroup$

Which technological path seems most promising to produce a quantum processor with a greater quantum volume (preferring fewer errors per qubit over more qubits), than Majorana fermions?

The preferred format for the answer would be similar to:

"Group ABC's method DEF has demonstrated better QV than using MF; as proven independently in paper G on page x, paper H on page y, and paper I on page z".

On Majorana fermions Landry Bretheau says:

These particles could be the elementary brick of topological quantum computers, with very strong protection against errors. Our work is an initial step in this direction.


Example of an insufficient (but interesting) answer:

In their paper "Robust quantum metrological schemes based on protection of quantum Fisher information", Xiao-Ming Lu, Sixia Yu, and C.H. Oh construct a family of $2t+1$ qubits metrological schemes being immune to $t$-qubit errors after the signal sensing. In comparison at least five qubits are required for correcting arbitrary 1-qubit errors in standard quantum error correction.

[Note: This theory of robust metrological schemes preserves the quantum Fisher information instead of the quantum states themselves against noise. That results in a good effective volume if they can construct a device utilizing their techniques and show that it scales.

While that might seem like one promising answer it's a single link (without multiple concurring sources) and there's no device built to show scalability. A low qubit device that's error free and unscalable or a device with many error-prone qubits has a low volume (and thus is "Not An Answer").]


Additional references:

Paper explaining Quantum Volume.

Qubits vs. Error Rate

After doing some research it looks like Graphene sandwiched between superconductors to produce Majorana fermions is the leading edge - is there something better? ["better" means currently possible, not theoretically possible or ridiculously expensive]. The graphic illustrates that over a hundred qubits with less 0.0001 error rate is wonderful, lesser answers are acceptable.

$\endgroup$
  • $\begingroup$ In general, it is best if your question can be explained without relying on illustrations. (even if there is text on the illustration) So, if you could add a summary of the relevant parts of your illustration, this question can be improved. Thanks! $\endgroup$ – Discrete lizard Mar 21 '18 at 11:57
  • $\begingroup$ this has no objective (undisputed) answer at the moment. It might be better to rephrase the question, for example asking how different architectures compare in terms of error tolerance, instead of which architecture is "the best" or "leading edge". $\endgroup$ – glS Mar 21 '18 at 18:38
  • $\begingroup$ @Rob that's not a good type of question. It's called a list question, and they are generally frowned upon. Unless you provide a way to evaluate which of two answers is the better one, the question doesn't help the site grow $\endgroup$ – ItamarG3 Mar 21 '18 at 19:57
  • $\begingroup$ +1'd the question because it sounds interesting, but just to note it, the presentation comes off as a tad defensive. Usually it's helpful to provide links on less common terms to make a post more accessible to readers, though there's no need to provide commentary on one's presumption of a "casual reader" vs. an "expert". Additionally I'd recommend constructing questions with a friendlier tone; this is, you're asking someone to provide information, so it's more appropriate to point out what you're looking for rather than specifying a filter for what'd be an "insufficient" answer. $\endgroup$ – Nat Mar 21 '18 at 23:09
9
$\begingroup$

That is indeed the most important question at the moment!

Superconducting qubits currently have the biggest devices. But will they continue to scale? Will short coherence times make it too hard for error correction to keep up?

Trapped ions are not far behind. But they have their own scalability issues.

Spin qubits should be great for scaling once they get going. They are still down in the few qubits at the moment, though.

Majoranas also are suspected to have some nice properties. But I’d have to see a single qubit before I declare them to be the leading edge.

Photonics are also a viable strategy. In fact, the first cloud based quantum device was photonic. A few startups are also based around photonic based approaches, such as the one described here.

$\endgroup$
  • $\begingroup$ Thanks for your interest in the question and answers. You've offered three suggestions with one or two links each. Can you explain and use a few (for each proposed, better technology) highly reputable links (like the ones provided) supporting your position that these have a lower error rate with higher qubits. $\endgroup$ – Rob Mar 21 '18 at 12:40
  • $\begingroup$ Some people, like those at Xanadu, might be disappointed with seeing photonics technologies being left out! $\endgroup$ – glS Mar 21 '18 at 18:47
  • $\begingroup$ @gIS That's true. I was visiting Bristol a few weeks ago, and so should really know better. $\endgroup$ – James Wootton Mar 21 '18 at 18:57
  • 2
    $\begingroup$ @Rob I wouldn't want to dive any deeper into any. The question as asked is one with as many possible answers as there are quantum information theorists. $\endgroup$ – James Wootton Mar 21 '18 at 19:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.