Chance of building practical quantum computers

[mok-kong shen]

Wiki on quantum tomography says: "The number of experimental configurations (state preparations and measurements) required for quantum process tomography grows exponentially with the number of constituent particles of a system. Consequently, in general, QPT is an impossible task for large-scale systems." Doesn't this fairly clearly indicate that the chance of building a moderately sized quantum computer is extremely low from the very beginning, since verification of computer hardware is necessary in it's design, manufacture and maintenance?

 

[BvU]

Check http://www.tnw.tudelft.nl/nl/over-faculteit/afdelingen/quantum-nanoscience/medewerkers/onderzoeksgroepen/quantum-transport/research/background-information/quantum-computation/[/URL] or [PLAIN]http://www.tudelft.nl/en/current/latest-news/article/detail/einsteins-ongelijk-delfts-experiment-beeindigt-80-jaar-oude-discussie/[/URL]

These guys show some nice results [URL]http://www.nature.com/nature/journal/vaop/ncurrent/full/nature15759.html[/URL] and even in the times [URL='http://www.nytimes.com/2015/10/22/science/quantum-theory-experiment-said-to-prove-spooky-interactions.html']here[/URL]

 

[Strilanc]

My computer has 8 GiB of memory. The number of states it can be in is so large that it takes a billion digits to describe it. No one will ever ever have the time to validate that each of those states works. And yet it does work, mostly.

The trick is to not attack the state space as a black box. The state is made up of repeated pieces interacting in common ways. Failures tend to break huge swaths of the space, instead of just a single state. And even if a single state was failing somehow, it's probably hard for the user to hit that state.

The same thing applies to quantum computers. Validate the pieces. Do statistical tests on the whole. Rely on truly subtle problems being hard to hit in practice. Understand your error model and use it to guide testing. If users do find that an algorithm consistently triggers a problem, include that algorithm in your test suite. Be good enough instead of perfect.

John Martinis recently wrote a paper on basically this subject, though at a smaller scale: Qubit metrology for building a fault-tolerant quantum computer.

 

[mok-kong shen]

If I don'r err, Martinis' paper doesn't mention quantum tomography. Hence my layman's questions: (1) Could quantum tomography play at least some role (i.e. even if it plays a comparatively smaller one than quantum error correction) in affecting the issue of the possibility of practical realization of quantum computing in the future (since the hardware needs diverse sorts of verifications)? (2) What is the relationship between quantum error correction and quantum tomography?