Experts Hail Quantum Computer Memory Stability Breakthrough

cold fjord writes "The BBC reports, 'A fragile quantum memory state has been held stable at room temperature for a "world record" 39 minutes — overcoming a key barrier to ultrafast computers. 'Qubits' of information encoded in a silicon system persisted for almost 100 times longer than ever before. ... "This opens the possibility of truly long-term storage of quantum information at room temperature," said Prof Thewalt ... unofficially, the previous best for a solid state system was 25 seconds at room temperature, or three minutes under cryogenic conditions. ... What's more, they found they could manipulate the qubits as the temperature of the system rose and fell back towards absolute zero. At cryogenic temperatures, their quantum memory system remained coherent for three hours. "Having such robust, as well as long-lived, qubits could prove very helpful for anyone trying to build a quantum computer," said co-author Stephanie Simmons of Oxford University's department of materials. ... "We've managed to identify a system that seems to have basically no noise." However she cautions there are still many hurdles to overcome before large-scale quantum computations can be performed. ... "This result represents an important step towards realizing quantum devices," said David Awschalom, professor in Spintronics and Quantum Information, at the University of Chicago. "However, a number of intriguing challenges still remain." — Abstract for the paywalled academic paper."

Re:They will break all the encryption

[harvestsun]

That's probably the most common misconception about quantum computers... They do not solve all NP problems efficiently, only a subset called BQP: "bounded error, quantum, polynomial time". It is believed that this is completely disjoint from the "NP complete" subset.

Re:They will break all the encryption

[sd4f]

I actually went to a talk by a Prof. Michelle Simmons on this last night, and asked that question. My understanding is that it would just does all the calculations at once, in a massively parallel operation (which obviously isn't efficient). I'm no computer scientist (just a mechanical and mechatronic engineer) and I don't really know anything behind quantum mechanics, but the other thing mentioned in the talk about a quantum computer is that it would have perfect security (her words), because, and now i'm relying on memory, a quantum computer doesn't store data like a classic computer, as it can't be perfectly replicated, so the quantum computer needs to keep the qubits active for as long as possible in the computer (hence the importance of the coherence time, as stated in the article). Because a quantum computer is an adiabatic system, it sends the 'energy' from one place to another. Eavesdropping would mean you reroute that energy, and it doesn't go to its intended place.

A lot of it went over my head, so take this with a spoon of salt, as I could have botched it up, but that's the gist of my understanding, and off-topic info

Re:

[HuguesT]

Unless P=NP of course, in which case they are all in the same class.

Re:

[JoshuaZ]

Actually, this does not follow either. Right now it is open if BQP (the set of decision problems solveable by a quantum computer in polynomial time http://en.wikipedia.org/wiki/BQP [wikipedia.org]) lives in the polynomial hierarchy http://en.wikipedia.org/wiki/PH_(complexity) [wikipedia.org]. Most complexity classes we care about live between P and NP, but it turns out that PH (a class believed to be much larger than even NP) will also collapse to P if P=NP, so there are larger complexity classes that also collapse if P=NP. However, it is

Re:They will break all the encryption

[JoshuaZ]

The second they perfect this, they will be able to try all the keys at once and the right one will be solved instantly. All of our current generation of encryption relies on n-p complete algorithms and will become worthless

There is a lot wrong with this.

First of all, quantum computers cannot as far as we know solve NP-complete problems efficiently. There's no known way for that to happen, and most experts expect they cannot. What they can do, is solve specific classes of problems more efficiently than classical computers. The most obvious example of this is factoring integers, which seems to be very difficult for a classical machine, but which can be done quickly by a quantum computer using Shor's algorithm. http://en.wikipedia.org/wiki/Shor's_algorithm [wikipedia.org]

Second, none of these algorithms are instantaneous or even remotely so, but rather scale wit the size of the input, generally with a polynomial.

Third, while many forms of crypto would be vulnerable, including RSA and elliptic curve cryptography, not all forms of cryptography have known vulnerabilities. This connects with the earlier issue of NP completeness. No form of crypto relies at this point on an NP complete problem. Factoring for example is in NP (which means roughly that one can easily convince someone that one actually has the factorization), but it is likely not NP complete. A problem is NP complete if (roughly speaking) it lives in NP, and if you have access to a black box that solves the problem then you can solve all NP problems. At this point, factoring is strongly suspected not to be NP complete, because that would lead to the collapse of the polynomial hierarchy http://en.wikipedia.org/wiki/Polynomial_hierarchy [wikipedia.org], which is strongly conjectured to occur.

Quantum computers will likely have real world consequences if we ever get them to work on a large scale, and some of those consequences will be cryptographic. But thinking that they'll somehow blow up all cryptography is just hype. If you want to read a good book, without hype, that does a good job of explaining how quantum computing actually works, I recommend "Quantum Computing Since Democritus" by Scott Aaaronson. The book does assume some slight comfort with linear algebra, a very tiny amount of group theory, and some calculus, but that's it. It is aimed at people in technical fields other than quantum computing to understand what it is about. It is highly readable, and I strongly recommend it.

Re:They will break all the encryption

[JoshuaZ]

Er, "strongly conjectured to occur" should be "strongly conjectured to *not occur*. Need to proofread more when using preview. Also, as long as I'm replying to myself, note that the set of problems which can be solved easily by a quantum computer is BQP http://en.wikipedia.org/wiki/BQP [wikipedia.org], but it is likely that for many practical applications, one will want the set of problems where a quantum computer can quickly convince a classical computer that it really has a solution, and this set, ZBQP, is likely much much smaller https://complexityzoo.uwaterloo.ca/Complexity_Zoo:Z#zbqp [uwaterloo.ca]. Factoring lives in this smaller set because the classical computer can check the quantum computer's work essentially by just multiplying together the factors given by the quantum machine.

Re:

[JoshuaZ]

So, the result for encryption may not be that large, but that's by far not the only thing that people want to use quantum computers for. They can be used to optimize a number of problems that show up in real world conditions, and do quantum simulations for chemistry where actually observing the reactions in detail would be too difficult.

Re:

[HuguesT]

Yes, quantum computer ought to be really efficient at simulating quantum phenomena. This is really useful, because simulating them on classical computers is very inefficient, and so our understanding of many physical processes should improve.

Re:

[Anonymous Coward]

Time to buy that potato farm in Idaho...

PROTIP: Use your Bitcoin to buy the farm.

Re:

[fuzzyfuzzyfungus]

More likely, there would suddenly be a huge demand for unbreakable quantum encryption, followed by massive investment in developing quantum computing technologies.

Unless they mean something rather different by 'quantum encryption' than the present usage, it won't be of much use.

If you are particularly paranoid, and operate on the theory that your fiber isn't being tapped, quantum encryption comes in at a price that compares favorably to having trusted guys with guns stand around keeping people away from your fiber. If, however, you don't have the luxury of a continuous run between you and the destination (like, oh, almost everybody), the fact that a third party ha

Re:

[fuzzyfuzzyfungus]

That was my intended point: very few entities actually have a quantum channel between point A and point B. Virtually everybody, even the ones with fiber links, has a link to some local telco or peering point and then is N hops away from their destination. At present (and barring substantial advances in all-photonic switching), that means that a 'compromised' quantum channel is the expected state, since there are multiple hops between you and your destination.

If you have the cash for old-school point-to-p

Re:

[TheTrueScotsman]

Symmetric (shared-key) algorithms do have vulnerabilities to quantum computing: Grover's algorithm allows a brute-force search to be performed in square-root time of the key size (i.e. 256-bits -> 128-bits, 128-bits -> 64-bits). This isn't a big problem as long as you're using at least 256-bits.

Re:Nice, but...

[Ashenkase]

I fail to see how KFC can leverage this technology... unless of course... they can make the elusive "n" down sandmich. O now that is ebil.

Re:

[account_deleted]

Comment removed based on user account deletion

Re:

[cold fjord]

I think another chain beat them to it. Isn't there an Arby-Qbit [youtube.com] sandwich?

Re:

[Joining Yet Again]

I would assume that the NSA is at least a decade ahead of open academia on all problems relevant to them.

Re:Nice, but...

[cold fjord]

I would guess 15-20, more or less, depending on the specific application. The history of the NSA's involvement with the DES encryption algorithm is instructive.

NSA's involvement in the design [wikipedia.org] (of DES)

I would also highlight the last sentence in the section: " Bruce Schneier observed that "It took the academic community two decades to figure out that the NSA 'tweaks' actually improved the security of DES."[12] "

Re:

[Anonymous Coward]

I am not sure about such assessments. It's known that the NSA and GCHQ hit on public key cryptography just a few years before academia did, in the 70s. However, at the time cryptography research in academia was not all that widespread, at least not in the west, probably in part because of pressure from said agencies. Analogously with DES. Likewise, there are strong indications that the development of ECC caught the NSA by surprise. These agencies no doubt have a far deeper hands-on experience that researc

Re:

[cold fjord]

Hard to say. Anyone can come up with a new idea, but only the spooks get to see everything they have plus what academia produces. As the biggest employer of mathematicians they have the potential to take a new idea and explore or develop it quickly. But it is also the case that math breakthroughs can hinge on the breakthrough insights of a single gifted (or gifted and dogged) individual. It is also the case that the details of a particular technology matter. If you compare the gross scheme of the Enigm

Re:

[mikael]

A couple of decades ago, even discussions on the factorisation of integers would be discouraged by the FBI, and exports of software that did 128-bit encryption and above required an official federal application for an export license.

Re:

[cold fjord]

I think you made a great post. As Bruce notes in that piece, nobody knows. You also can't overlook the possibility that they hold back on vulnerabilities until an exploit for them is in reach of the public or some other adversary. It is also worth noting that working on this year's NIST standard isn't all they do. There are about 200 countries out there, and plenty of different encryption schemes to keep them busy cracking. They also have their own US government codes design, test, and vet. There is

And it was so simple to do...

[TempeNerd]

I am just amazed at the technology that is going into making this new qubit.

First off it is "...built with a highly purified form of silicon" and one qubit requires "... the spins of the 10 billion or so phosphorus ions..."

Now THAT is engineering!

Re:

[TheSync]

The problem is that I don't think anyone has ever built a quantum gate with phosphorus doped silicon...

Re:

[account_deleted]

Comment removed based on user account deletion

Another observation

[Okian Warrior]

We are also testing the boundaries of the physical universe in a completely new realm.

The number of states in a quantum-entangled set of particles goes exponentially with the number of particles. For 10 particles (entangled) it takes 2^10 states for the universe to represent the possible outcomes. For 1,000 entangled particles, the number of states is 2^1000.

The number of particles in the entire universe is only about 2^80.

Managing 1,000 entangled particles would require the universe to keep track of a staggering amount of information. Does the underlying machinery have information-space this big? No one knows.

For the first time, we can measure the boundaries of the physical mechanism that underlies the universe in a completely different realm: information capacity. This is analogous to a program probing the limits of RAM memory by seeing how much it can allocate.

Here's hoping that we don't find a buffer overflow.

Wrong number of atoms in the Universe

[Myria]

It's about 10^80 [wikipedia.org], not 2^80.

I think we'll find that the amount of energy required to hold X entangled particles in coherence will be exponential in X. This would make quantum computing essentially worthless.

If not, wake me when we get to 2048 qubits, for the original Xbox's public key and I have some unfinished business from last decade...

Good catch

[Okian Warrior]

It's about 10^80 [wikipedia.org], not 2^80.

Good catch. I misremembered the number from my original source, which also lists it as 10^80.

Re:

[cold fjord]

Managing 1,000 entangled particles would require the universe to keep track of a staggering amount of information. Does the underlying machinery have information-space this big? No one knows.

Quantum entanglement appears to be a key element of photosynthesis [lbl.gov], and systems of more than a hundred million entangled photons have been achieved [nature.com].

Windows

[TVDinner]

39 minutes.....that's long enough to run Windows between BSODs

What a load of crap

[TomGreenhaw]

Quantum entanglement - really??? Can a cat be alive and dead at the same time? Huh???
Could all this be non-science and a misinterpretation of statistical math that doesn't really describe reality?
The underpinning of quantum computing is the idea that a qubit has the ability to be in more than one state simultaneously. This is simply illogical.
Don't shrug it off and say "things work counter-intuitively at the subatomic level"; that's a philosophical cop out and is not science.
The quantum is a profound re

Re:

[JoshuaZ]

Your point that the word "quantum" originally referred to the discretized nature of energy states is true but unhelpful. Words have distinct meanings apart from their etymology. Don't think that because scientists started using a word because of it having one Greek root doesn't mean they cannot use the word in a more general context when they are clear about what they mean.

The underpinning of quantum computing is the idea that a qubit has the ability to be in more than one state simultaneously. This is simply illogical. Don't shrug it off and say "things work counter-intuitively at the subatomic level"; that's a philosophical cop out and is not science.

Actually, if anything, this is a biological/psychological statement. You are confusing logic with intuition. Humans evolved on the mediu

Re:

[TomGreenhaw]

"if you agree that these statistics do describe accurately what results to get"
Yep, I agree completely and that's why I say that statistical formulas can be useful for large numbers of entities. But for a single particle - nope - not the right tool. And herein lies the core the argument - Bohr vs. Einstein
My position is that Einstein was right, God does not play dice and the universe is fully deterministic. Unfortunately the mathematical model to describe a deterministic universe would need to take into c

Re:

[JoshuaZ]

But we have gotten quantum mechanics to work well on single particles. There have been many experiments involving individual electrons or individual photons (fewer with photons since it is very difficult to send out a single photon). Moreover, this actually misses what quantum computers rely on: they aren't relying on the behavior of the individual particle as much as on the entangled state, exactly where you seem to think that the statistics works well. You may also want to look into Boson Sampling http: [scottaaronson.com]

Re:

[TomGreenhaw]

You make two good points.
"That you can "get away" with something isn't a reason to do it." I made a poor choice of words. The experiment described in TFA is actually very cool (pun intended). What I'm trying to say is that the people who have a lot at stake in this field are afraid to rock the boat by expressing their misgivings about some of the underpinnings of quantum computing. I'm not trying to win a nobel prize here or yell fire in a theater, my aim is to gain a better understanding of reality and s

Re:

[JoshuaZ]

On the contrary, if you talk to the people who are doing this, many of them acknowledge that something might not work, and they are eager for that because it would be a glimpse of new physics. At this point, we know that the Standard Model isn't all there is, but we don't know exactly where it breaks down. If we can set up what should be fully functioning quantum computers and then they don't do what we want, that will be Nobel Prize level discoveries by itself. As to your last bit, I don't know what you me