Breakthrough for Quantum Measurement

said_captain_said_wo writes to tell us that PhysicsWeb is reporting that two teams of physicists have developed a new method for measuring the state of quantum bits in a quantum computer without disturbing the state. From the article: "In the future, the Josephson capacitance could be used for operations in a large-scale quantum computer," says Mika Sillanpaa of Helsinki University. "The Josephson inductance and Josephson capacitance together would also allow us to build new types of quantum 'band engineered' electronic devices, such as low-noise parametric amplifiers."

Comments

Is quantum computing useful beyond decryption?

[jimmyhat3939]

The problem with quantum computing, as I understand it, is there are very very few applications.

Essentially, it's only useful in a situation where you need to repeatedly run the same computation over and over again with different input values to see which of those values produces a valid output.

I have a friend who has suggested repeatedly that eventually computers will contain some sort of quantum processor that helps with such tasks as gaming. I don't think this is realistic because of the serialness of the tasks that quantum computing tackles. In particular, something like rendering an environment in real-time won't be helped because there's an unpredictable input (the human).
--
Free 411! 1-800-411-SAVE

Implications in reverse order

[Flying pig]

Talk about looking for grant funding! Problem is, scientific illiterates in Government etc. think they understand what a quantum computer can do (application a long way in the future if at all) but not what you can do with very low noise parametric amplifiers (which might be relatively near term applications.) In terms of exciting progress in studies of brain function, small scale biochemistry, remote sensing and signal processing, very low noise amplifiers are critical components, whereas quantum computers don't yet exist, and by the time they do conventional computers should be adequate to deal with a lot of the data processing.

Not to knock the discovery, which is very interesting, but it's a pity quantum computers have to be dragged into everything to justify research. I doubt that Tom's Hardware will be reviewing millikelvin coolers for your qubit box any time in the next 20 years (though I'd like to be proved wrong)

spooky action at a distance

[ericcantona]

presumably, given entanglement, measurement of qbit state allows potentially for instant communication ? (which would be really spooky!).

Unchanged State

[squoozer]

I thought the state had to be changed to measure it or am I confusing a technique used in quantum crytography with this technique in quantum computing. As an ex-chemist my understanding of things quantum was never that good anyway but I seem to remember someone saying that in order to measure something you had to change it. Any physicists in the house?

Physicist in the House

[anomalousman]

They can do what they say, but it's a lot more trivial than measuring the entire quantum state of the system, which is, as others have suggested above, impossible.

The Heisenberg Unccertainty principle implies that measuring a quantity must add noise in the conjugate quantity. For example, measuring the momentum of an object spreads out the wavefunction. Another example, measuring the state of a qubit (whether it is a zero or a one) destroys the relative phase between the zero and the one.

So the "non-destructive" measurement they are talking about means that they aren't changing it from a zero to a one or vice-versa. But they are (and must) destroy the information about the phase of the qubit state during the measurement. For a more in-depth discussion, look up "quantum nondemolition measurements".

mis-statement (I think)

[sdedeo]

It's been awhile; I do GR now, not QM (much simpler.) But any measurement will change the state; this is the famous "collapse of the wavefunction" (in the Copenhagen interpretation.) What they mean is that the measurement will collapse the wavefunction as usual, but that it will not then alter the system being measured so that the state changes. i.e., if the amplitude is 0.1 A and 0.9 B, and the measurement collapses the wf to B 90% of the time as it will, then when the measurement is done the system will be B 90% of the time as expected, and it will be B "the right" times.

Quantum cryptographic links?

[evilviper]

Sounds to me like the security of quantum fiber-optic links are now in question. This isn't directly applicable to taping one, but it's a start.

(Not a quantum physicist, but I can play one on slashdot can't I?)

Re:Quantum cryptographic links?

[Anonymous Coward]

This is a technological breakthrough, not a theoretical one. It allows a practical measurement to be closer to an ideal quantum measurement. In quantum cryptography it is always supposed that the eavesdropper may do any perfect quantum measurement, even though this is not technologically feasable today.

For a measurement of a totally unknown state, the state is always changed by the measurement. But if you know that the qubit is in one of the basis states it is not changed by an ideal measurement. Now we're closer to doing this in practice.

Re:Shroedinger's cat?

[Anonymous Coward]

Does this mean that we can find out if the cat is dead without opening the box? Sure sounds like it.

That's exatcly what it means, the way the headline presents it, which would mean that QP is disproved, which again leads me to beleive that the poster has misunderstood something.

Re:Shroedinger's cat?

[tendays]

What happens if an ant crawls into the box for example? Because it's not 'really' an observer the cat is still half alive??

Yes the cat is still half alive, and the ant is half seeing a dead cat and half seeing a cat alive. What happens when the ant walks in the box is that its state gets correlated to that of the cat.

Note that the either state of the ant is unaware of the other.

When you open the box you will either see a dead cat and an ant that has been seeing a dead cat all along, or a living cat and an ant that has been seeing a cat alive all along

Oh, and finally, there are many indirect but observable consequences of these superposition states, and they are precisely what quantum computing takes advantage of.

by the way, sorry I know this is off topic, but is anyone else having problems logging in in slashdot.org subdomains? That problem started for me a couple of weeks ago : I'm logged in slashdot.org but not in science.slashdot.org or it.slashdot.org etc so I can't mod in these domains unless I comment first

Shor's Algorithm

[stelmach]

I'm trying to grasp what the implications are of this. Let's take Shor's algorithm as an example. It is my understanding that the Quantum Fourier Transform (QFT) is applied to the result of the algorithm to peak the probability amplitudes, which will help the result to collapse into the correct state when measured. So does this mean that the QFT will not need to be applied, and the result of Shor's algorithm can be read with 100% accuracy?

A Golden Age is Coming

[TheZorch]

Quantum Computers will usher in a golden age in computing. There are all sorts of applications that they could be used for. For a time they'll serve a role that most super-computers today serve and that's for engineering computations and scientific experiments that require massive number crunching.

For long term space travel like the proposed mission to Mars a quantum computer would be invaluable. It would be able to monitor the crew and spacecraft faster than today's computers and will be able to react to any kind of critical issue 100x faster. Please, no "2001/Hal" comments. I'm being serious.

Also, quantum computers could be used for gene sequencing that can be done in minutes rather than hours, months, or years for the creation of new drugs or gene therapy. A single quantum computer could be used to replace dozens of computers in a corporation's server room. Just one machine could do the work of 20 or more so you don't need a separate database server, email server, web server, web proxy server, or any other kind of server a large company would need. These computers would benefit businesses like Ford, GM, and all the other car makers allowing them to make better engineered cars.

I can also imagine the graphics industry would benefit. Imagine if Pixar had one of these machines. Imagine being able to render a movie at final-draft production quality in "realtime". We'd also finally have a computer that could make Virtual Reality better than it has been in the past. The applications for this technology aren't as limited as you might think.

Useful?

[necro81]

From RTFA, I am wondering if this new discovery will actually be of much use to anyone. The apparatus involves cooling down to a few millikelvin. I am guessing that this is so that the thermal noise in the circuit is greatly reduced, and also because the superconducting threshold of whatever their Johnson capacitor is made from might in fact be that cold. Pure copper becomes a superconductor, but not until several degrees Kelvin, I believe.

In any event, cooling down to such temperatures implies a couple of things: lots and lots of very expensive equipment to cool down a tiny tiny volume of space. Even the first transistors didn't require such great lengths.

The article also makes reference to the capacitance of the Johnson capacitor changing signs depending on the state of the qubit, which is part of how the whole thing works. Does this mean that someone has discovered negative capacitance? Whoa! What would that mean?

Re:Wrong

[tomhudson]

Um, no. You're wrong. The 2-slit experiments I'm referring to neither require nor use a mirror. You're referring to something else.

As for the temperature, there is no way that a thermometer of any size can't affect the sample its measuring unless it already is at the same temperature as the sample.

Every measurement, even in the macro universe, affects the thing measured. That this would NOT be the case at the quantum level, while counterintuitive at first, would upon reflection be surprising.

Also, there is no need to resort to a superposition of states (which has always been a bit of a brainfuck imho) when there are better models. Superposition is only required if the time scale is unidirectional and can't be "rotated out" of the question and replaced by another vector, which has never been shown to be the case. The copenhagen gang lacked sufficient imagination to see the obvious.

But that's just my opinion, and this is slashdot, and its not Tuesday :-)