New Quantum Record: 14 Entangled Bits

Tx-0 writes "Quantum physicists from the University of Innsbruck have set another world record: They have achieved controlled entanglement of 14 quantum bits (qubits) and, thus, realized the largest quantum register that has ever been produced. With this experiment the scientists have not only come closer to the realization of a quantum computer but they also show surprising results for the quantum mechanical phenomenon of entanglement. By now the Innsbruck experimental physicists have succeeded in confining up to 64 particles in an ion trap. 'We are not able to entangle this high number of ions yet,' says Thomas Monz. 'However, our current findings provide us with a better understanding about the behavior of many entangled particles.' And this knowledge may soon enable them to entangle even more atoms."

Comments

Don't read this!

[Anonymous Coward]

Quantum first post!

"Superdecoherence"

[AdmiralXyz]

From TFA:

In addition, the physicists of the University of Innsbruck have found out that the decay rate of the atoms is not linear, as usually expected, but is proportional to the square of the number of the qubits. When several particles are entangled, the sensitivity of the system increases significantly.

This is somewhat troubling, isn't it? If the decay rate is quadratic in the number of qubits, and this turns out to be due to some fundamental physical law as opposed to limitations of the current technology, does that mean we can never have quantum computers with any significant amount of memory?

Re:"Superdecoherence"

[DWMorse]

does that mean we can never have quantum computers with any significant amount of memory?

16k ought to be enough for ANYbody.

Re:"Superdecoherence"

[Anonymous Coward]

Since for some algorithms the computational power is exponential in the amount of quantum memory, you can do "significant" stuff without a lot of memory.

Re:"Superdecoherence"

[greeneggs2000]

From TFA:

In addition, the physicists of the University of Innsbruck have found out that the decay rate of the atoms is not linear, as usually expected, but is proportional to the square of the number of the qubits. When several particles are entangled, the sensitivity of the system increases significantly.

This is somewhat troubling, isn't it? If the decay rate is quadratic in the number of qubits, and this turns out to be due to some fundamental physical law as opposed to limitations of the current technology, does that mean we can never have quantum computers with any significant amount of memory?

Not really. The researchers trapped and entangled 14 ions in a single ion trap. Quantum computers based on ion traps will have thousands of traps, with never more than one or two ions per trap. (Machines with hundreds of traps have been tested, ions moved between traps, etc.; see, e.g., [1]) It has been known since at least 1997 [2] that you can't have a scalable system with only a single ion trap (that would be true even were the decay rate quadratic in the number of ions per trap).

[1] Home, J. P. et al. Complete methods set for scalable ion trap quantum information
processing. Science 325, 1227–1230 (2009). arXiv:0907.1865 [quant-ph]
[2] Wineland, D.J. et al. Experimental issues in coherent quantum state manipulation
of trapped atomic ions. J. Res. Natl. Inst. Stand. Technol. 103, 259–328 (1998). arXiv:quant-ph/9710025

By the way, an arXiv link for this article is arXiv:1009.6126 [quant-ph].

Re:"Superdecoherence"

[Kjella]

Since for some algorithms the computational power is exponential in the amount of quantum memory, you can do "significant" stuff without a lot of memory.

Compared to the gigabytes of memory on your average computer, sure. 1 kilobyte = 8192 bits would be huge. But 14 bits? At most 2^14 = 16384 "classic" operations at once. I've never heard how many IOPS you'll get from a quantum computer but my impression is that you need many more qubits to beat a supercomputer.

Re:"Superdecoherence"

[Angst Badger]

Since for some algorithms the computational power is exponential in the amount of quantum memory, you can do "significant" stuff without a lot of memory.

That may be so, but I have a feeling that they'll still need to be able to implement at least three registers to accomplish anything, and they haven't quite made it up to being able to implement a single short int. The idea of quantum computing has a lot of potential, but so does holographic memory, and they've been promising results there since the 1960's. When you're fighting entropy on as many fronts as they are, there's good reason to be pessimistic.

Re:"Superdecoherence"

[frnic]

"Quantum computers based on ion traps will have thousands of traps"

Let me be the first to say, "Who could ever need more than thousands of ion traps..."

Re:"Superdecoherence"

[Anonymous Coward]

Well, doesn't reading the answer break the entanglement anyway? As far as I can tell, the full process of a quantum computation is to tangle up a fresh batch, do the computation (essentially trying every possible answer in the same instant), read it (collapses the state of the particles into the correct answer). So they only need to stay entangled for a very brief time. And the idea is to eventually get enough quantum bits together to do a Metric Crapload of computation all at once. In an analogy that is probably too oversimplified: One thread on a 2ghz system can theoretically do 2 billion operations per second; but if you can do the setup-compute-teardown of some quantum bits in one second, and you've got, say, 40 quantum bits, you can do a trillion operations per second.

I don't think there are any problems a 14 qbit computer could solve that a standard computer couldn't do in well under a second anyway. But even then, there may be some kind of case where you have a few million such problems you need to do, and a singly insignificant speed advantage dominates in bulk. For example, perhaps some obnoxiously compute intensive particle simulations can be decomposed into still-huge number of 14-bit NP problems, and by doing so you replace a warehouse sized cluster with one fridge-sized unit. Though again, the potential for this goes up drastically with the number of quantum bits.

Re:"Superdecoherence"

[InfoJunkie777]

You should get upgraded to at least TWO for FUNNY!

Re:"Superdecoherence"

[Urkki]

Calling qubits 'memory' is a little misleading. The number of qubits that can be entangled at once is more analogous to the number of bits per register/operation in a classical computer. You could have 100 registers of 14 qubits and each register would individually have the same decay rate proportional to 14^2. That's not to say that this isn't a big hurdle, but it's not as big of a hurdle as you're making it out to be.

Are you sure? If those 100 registers aren't entangled with each other, how could you combine them in any useful way? Wouldn't it be just same as having 1 register, and using it 100 times in a row, so not really any help in solving a problem, which is too large to be solved with a single register?

Re:"Superdecoherence"

[SoftwareArtist]

Someone please correct me if I'm wrong, but my understanding is that computing power depends on the number of entangled bits. The promise of quantum computers is that they can solve in O(N) time certain problems that a conventional computer would need O(exp(N)) time to solve - but only if all N bits are entangled. If you're limited to 16 entangled bits, you can't solve problems any larger than N=16 without losing the linear scaling.

Re:"Superdecoherence"

[oliverthered]

it depends if you can control the decay rater otherwise... physics is just a theory, until it's not then some apparent physical law doesn't mean much by the way of limitations.

Re:"Superdecoherence"

[uninformedLuddite]

does that mean we can never have quantum computers with any significant amount of memory?

16k ought to be enough for ANYbody.

Just how do you propose to know how much memory anyway?

Re:"Superdecoherence"

[ZSpade]

That's what arrays are for, no?

14 quantum bits

[kvvbassboy]

Hell yeah! In a few years, I will be able to play Super Mario on a quantum computer!

Re:14 quantum bits

[VortexCortex]

Hell yeah! In a few years, I will be able to play Super Mario on a quantum computer!

Yes, but then you'll have to deal with Bowser's Peach Paradox -- The game will start with the Princess being both captured and not captured, and you'll only find out which if you complete the game and observe the ending.

"I'm sorry Mario, but our Princess exists in a super position of both being in another castle, and awaiting your return safely at home."

Only after you observe the game's ending will you discover the game's plot:
You either attempted to save the Princess from the evil clutches of King Kupa,
or it's another case of Mario going mad and destroying an innocent kingdom for no good reason.

Of course the credits will either reveal that the game's events haven't taken place yet (it was all a dream (ala Mario 2), ), or that the story has all happened before, an infinite number of times, and the princes might have just been captured again!

Talk about replayability...
Insert Qubits to Contiue.

Re:14 quantum bits

[infurnus]

Only after you observe the game's ending will you discover the game's plot:
You either attempted to save the Princess from the evil clutches of King Kupa,
or it's another case of Mario going mad and destroying an innocent kingdom for no good reason.

So basically, the plot to Braid?

Re:14 quantum bits

[Plekto]

It won't run any faster, though. The distances in a typical computer chip are so tiny and there are enough choke-points in a typical computer that it really won't run any faster. I think that's what people are forgetting - that this isn't about speed or faster computers but about long-distance communication.

Now, being able to communicate with a person, say, on the moon, in real-time would be useful. Or a computer network between planets. Also, transmission from anyplace on the planet. Though that could easily be a problem - imagine trying to track down a suspect who is using a quantum based communicator.

Re:14 quantum bits

[the_other_chewey]

Quantum entaglement can not be used for faster-than-light transmission of information.

Re:14 quantum bits

[Plekto]

If one pair or group was physically moved to the moon (in theory), it would effectively become FTL communication though no actual data is being transmitted anyplace. The potential advantage of quantum entanglement is that you do an end-run around the entire problem of distance. You could have a device in theory 20 LY away and get data from it instantly. Of course, there's the issue of bandwidth and all, as well as numerous technical issues concerning longevity and repeatability.

My best guess for a potential use would be to send a large amount of data in a one-shot transmission as a notification or in an emergency. Kind of like those things hikers use so that people can find them when they are lost - just far better - press the panic button and it spits out your coordinates to the other device it's matched to. Even if you can't get a phone or GPS signal at your location. (ie - say, you're in a cave 200 ft below ground).

Re:14 quantum bits

[Dekker3D]

The plot to Braid, multiplied by Braid, with a little Braid thrown in.

I
CAN'T
WAIT!

Re:14 quantum bits

[Captain Segfault]

No.

You can't use separated entangled qubits to send information faster than light. It doesn't work that way. There are a bunch of tricks and operations you can do, but none of them result in the other end being able to distinguish a change of state.

Re:14 quantum bits

[Plekto]

Well, we'll see. Some physicists think that we will be able to distinguish a change of state eventually (though this may be a LONG time in the future) and use it to do exactly this. Some do not. I think that we will overcome this "limitation" some day and be able to use it like this, since observing changes is really a technological problem on our end(kind of like how people said we couldn't ever fly to the moon. We can, but it takes amounts of energy and technology that 100 years ago, even, they would have considered absurd.

Shoot, 50 years ago, we couldn't see atoms. Or planets around other stars. If it's just a problem of observing the changed states, it's going to be possible. Some day.

Re:14 quantum bits

[mug funky]

i take it recoupling the qubits once the parts are removed from each other is not possible?

if it were, you could use time-domain encoding - it doesn't matter that the information is random as such, just that something happens.

Re:14 quantum bits

[grcumb]

Hell yeah! In a few years, I will be able to play Super Mario on a quantum computer!

Yes, but then you'll have to deal with Bowser's Peach Paradox -- The game will start with the Princess being both captured and not captured, and you'll only find out which if you complete the game and observe the ending.

Downside: Quantum superposition allows every event in the game to occur simultaneously and in parallel to the others, so the game is -quite literally- over before you know it.

Re:14 quantum bits

[c0lo]

Some physicists think that we will be able to distinguish a change of state eventually [...]. Some do not.

Meanwhile, there is a group of physicist are in a superposition of the state thinking that FTL is and is not possible... they pertain to the class of String Theorists.

Paradoxically, the nature of their thinking state is totally opposite to quantum mechanics: any attempt to get an answer from their part will NOT result in a collapsing of their thinking state into one of the defined choices, but rather in setting the mind of the asking person into an indeterminate and fuzzy state (i.e. the "decoherence of the observer" effect).

Furthermore, in deep contrast with the normal quantum entanglement (on which the super-decoherence was observed), the above mentioned sub-system of String Theorist are believed as becoming more stable as the number of scientists in the group increases - in other words, a successful conversion of a new scientist to the group (will require an O(0) effort - i.e. constant, even if non-negligible) is most likely to result in a supra-linear increase in the stability of the so called "group coherence" and their capacity to influence the outside world.

Notes of caution for the young and adventurous - a short term exposure of an external observer may result in an assessment of the "thinking state" as being "incoherent", even if a longer period of observation will most likely note that the discourse and argumentation show patterns that are stable and that strongly resembles rationality and method. The external observer is warmly advised to refrain from searching for the "method behind the madness" under the risk of a fate worst than "living in the basement of their Mum" - see the reference to the "observer decoherence" effect above.

Re:14 quantum bits

[dominious]

It won't run any faster, though. The distances in a typical computer chip are so tiny and there are enough choke-points in a typical computer that it really won't run any faster. I think that's what people are forgetting - that this isn't about speed or faster computers but about long-distance communication.

Two words: Shor's algorithm

Re:14 quantum bits

[Plekto]

Heh. Then again, quantum physics may be a fancy kludge for something else. Similar to medieval astronomy and their horrendously complex calculations to make everything "fit". We just don't know yet. So far, attempts to unify all of the theories together have completely failed. Something is plainly wrong and needs to be thrown out or re-done (maybe all of it even).

I think that there is a way to "observe" such a change in state without actually observing it. We would need to be able to master gravity along the same lines as how we have essentially mastered electricity, though. That way we could passively look at how it's affecting the fabric of space around it and figure out the alignment of the atoms without actually influencing them in any way. I suspect that we may need hundreds or thousands of years more before we gain such technology, though.

Maybe some bright person will come up with a fancy kludge. I think we'll eventually do it. But not in our lifetimes, that's almost certain.

Re:14 quantum bits

[Plekto]

Yes, I know about that. The issue is that while the processor might be faster, nothing attached to it is, so it effectively is spinning its wheels waiting for the rest of the system. You'd need a whole new motherboard design, new peripherals, new memory, and so on that could keep up. One weak link in the chain and the speed gains largely evaporate.

Re:14 quantum bits

[dominious]

If you run Shor's algorithm on a quantum computer it will take polynomial time as opposed to exponential time on a classical computer for integer factorization. So yes, it is about speed and faster computers.

Re:14 quantum bits

[Plekto]

Technically, yes, but it's like having a 1000 hp car on L.A. freeways. In the end, nothing really goes a whole lot faster unless you were to redesign everything from the ground up to be able to operate at those speeds.

Re:Very cool

[Anonymous Coward]

So support life extension. Why so many geeks are against life extension but for all kinds of projects that will take centuries to pan out is beyond me. Why explore space if you can't explore time?

Exciting news!

[osu-neko]

At the rate advances in quantum computing are coming, with more and more bits available to be used, I should be able to see quantum computers as powerful as ENIAC before I die! Alas, baring major medical advances, I'm unlikely to see anything as powerful as a quantum TI-99/4A...

big deal

[Quiet_Desperation]

I can get more things entangled by just leaving a couple extension cords unattended for a few days.

Re:big deal

[jjohnson]

My wire-clothes-hanger closet computer is orders of magnitude more powerful than yours. I just need seed funding to commercialize it.

"Closer"

[betterunixthanunix]

Unless they have some theoretical method of scaling up their design, this does not really bring us "closer" to useful quantum computing. In fact, TFA casts some doubt on scalability:

In addition, the physicists of the University of Innsbruck have found out that the decay rate of the atoms is not linear, as usually expected, but is proportional to the square of the number of the qubits.

Re:Very cool

[MrKane]

If you can travel close to the speed of light you can explore quite a bit of space, in a normal amount of time. You would also have travelled forward in time at a faster rate than your home world, and so would infact be exploring time, should you decide to return to your point of origin :)

Re:Very cool

[TheLink]

So how rich would you need to be in order to have a good chance of successfully doing that?

Seems like it would be more expensive and difficult to build a spacecraft with enough shielding, payload, the ability to go to near C for a number of years, turn around and return (at near C) intact, than it would be just to solve this pesky quantum computing problem :).

Yeah yeah, right...

[Robert Zenz]

...cut the chit-chat...does Linux run on it yet? scnr

Re:Yeah yeah, right...

[lennier]

...cut the chit-chat...does Linux run on it yet?

Yes and no.

Expect 20 in 20 yrs, 25 in 100 yrs, 30 in 1000 yrs

[gweihir]

Seriously, the speed the number of entangled quantums is rising with, clearly points to exponential increase in complexity. This means we will likely never see quantum computers that can be used for any real problem size. Not that this has been clear for about a decade or so.

Re:Very cool

[SuricouRaven]

You could skip the 'turn around and return' part. Head off as 99.99999C, very slowly circle around back to your point of origin. You'll be going too fast to stop, but hopefully in the intervening millenia someone will have worked out a way to decelerate you as a historical curiosity.

One more thing I won't be getting...

[WaffleMonster]

Sigh...no hoverboards, flying cars, mr fusion or quantum computers :(

It figures..that there would be no free lunch...all of the initial rants about instantly factoring huge numbers, solving impossibly complex problems have unsurprisingly turned out to be false.

If you can't scale the number of qbits in a single coherent system QCs are doomed.. All of this talk of linking separatly entangled systems to produce more powerful QCs is crap. If you don't get anywhere near expontential scaling as a function of qbits then game over.

Re:One more thing I won't be getting...

[Patch86]

All of this talk of linking separatly entangled systems to produce more powerful QCs is crap. If you don't get anywhere near expontential scaling as a function of qbits then game over.

Why?

Serious question. I honestly don't know much about q-computing, but batteries of small systems to make a single large system sounds pretty par for the course, technologically. Doesn't sound like there's any fundamental problem with that.

Re:One more thing I won't be getting...

[Anonymous Coward]

Why?

Serious question. I honestly don't know much about q-computing, but batteries of small systems to make a single large system sounds pretty par for the course, technologically. Doesn't sound like there's any fundamental problem with that.

There is nothing wrong with QC.. I have no doubt it will be useful in the real world at some point in the future.

The problem for me is that while yes you can always throw more CPUs at a problem there are whole classes of problems where this is infeasable. This was the whole point of QCs... To solve problems that were **impossible** to solve with "classic" computers.

  A massive supercomputer may be a several million times more powerful than my desktop and that makes it more capable and very useful...yes..but even millions of such computers could not come close to breaking a single private key before our sun runs out of fuel and becomes a white dwarf.

You can't solve problems requiring every atom on earth or for that matter a thousand universes to be a transister in a massive supercomputer to solve. Without n^qbit scaling the origional promise of addressing impossible dreams behind QCs will never really be realized.

Re:One more thing I won't be getting...

[Wandering Idiot]

#2 Scientists (*cough* physicists) would stop being so dang dogmatic.

The first part was so vague and general as to be useless, but for this part I'll bite: Go ahead, tell us about your perpetual motion machine/reformulation of General Relativity/antigravity machine/theory of everthing that the mean old dogmatic scientists refuse to believe. You are of course the first person in history to come up with such a thing and assume no one takes you seriously only because of small-mindedness rather than the inherent flaws and mistaken assumptions of your model. Or it may be a more general objection, and you're just mad at them for not believing in ESP and ghosts.

Re:Very cool

[cgenman]

Because life is not an int that can just be increased? "Life Extension" is largely a sham promoted by con artists?

It's not like most geeks are against medicine, living well, reducing unnecessary risks, etc. But to say that you're researching "Life Extension" is like saying that you're researching "engineering bigger things" or "making fast stuff." Sure, someone doing research biology into the breakdown of DNA over time can be said to be doing "life extension." But anyone who says that they're doing life extension is probably an Israeli microcorp that releases 6 months of press releases, then disappears leaving nothing behind but bewildered investors and a badly dated looking website.

Re:Very cool

[ComaVN]

hopefully in the intervening millenia someone will have worked out a way to decelerate you as a historical delicacy.

FTFY

First command!

[mfnickster]

The first command they'll run on the quantum computer:

cat schroedinger.txt | tee alive.txt dead.txt

Re:Very cool

[cavreader]

I might not see any practical results from this research but I am just glad there are people who work on projects such as this. That includes both the intelligent scientists, engineerss, and the people who fund this type of research knowing themselves that they may never see any results in their lifetime either but continue to support this type of research anyway.

about freaking time

[epine]

In addition, the physicists of the University of Innsbruck have found out that the decay rate of the atoms is not linear, as usually expected, but is proportional to the square of the number of the qubits. When several particles are entangled, the sensitivity of the system increases significantly.

I've long said that I wouldn't take quantum computing seriously until I saw an equation depicting a scaling bound. That day finally dawns a decade into the hype cycle. Amazing. Seriously, following the field is like studying optics without knowing the difference between lumens and lux. What kind of physical system has no bounding process?

This is the first such equation I've seen, but they don't indicate the base decay rate, or how many qubits it would take before the decay rate is unmanageable.

Furthermore, they don't indicate the stacking rate: how long it takes to entangle qubits as a function of N. There's got to be some value where the stacking rate and the decay rate interest. I'd like to know what that value is, with present approaches, and viable future approaches.

Now if only the media could keep becquerels, sieverts and coulombs per kilogram straight. The book could be titled "Lumens and lux for people who don't wish to remain dummies" and any colour other than yellow and black.

New world record?

[stor]

No fair! They changed the outcome by measuring it!

-Stor

Re:Very cool

[Joce640k]

You wouldn't even need to go in a circle. Mankind will have populated the entire galaxy by the time you drop out of light speed.

By this same logic we can also know for certain that traveling backwards in time is impossible. If it were possible then the first time machine would have been 'invented' shortly after the big bang, ie. somebody would use their machine to travel back there so they could claim to be the inventor of time travel.

Re:Very cool

[williamfrench4]

Is anyone actually against life extension, or do they just assume it's impractical?

Re:Nullo bits?

[mug funky]

god, i read every word of that.

fuck you.

Z-80

[haapi]

They couldn't have entangled a Z-80's worth of bits and called it good. Sigh.

Re:Z-80

[c0lo]

They couldn't have entangled a Z-80's worth of bits and called it good. Sigh.

Huh? Z80 was an 8-bit processor. Granted, it had more registries than this one (if I remember well, it actually had a pair of registry sets).

Re:native americans presidential bid; no fake weat

[mug funky]

less coherent than timecube.

So...

[G33kDragon]

You're saying it's 14 of these: http://www.guessmyimage.com/puzzle/silly-rabbit-trig-is-for-kids

Re:Very cool

[uninformedLuddite]

Define 'quite a bit'. IMHO you could explore Sweet FA in one of your normal human lifespans. Thank Cthulhu that I'm not from around here.

Imagine someday buying disks in pairs....

[Desmoden]

Where every bit on one drive is entangled with the other.

Move drive A any distance from drive B, data remains in sync. ....errr...or you buy one drive and the Gov has the other...hmmm....

Re:Expect 20 in 20 yrs, 25 in 100 yrs, 30 in 1000

[Interoperable]

The trouble is that no truly scalable proposal for QC has been developed yet. The hope would be that once a suitable system was found, it wouldn't be exponentially more difficult to add qbits. Photonic qbits have very different problems from trapped ions, for example. Not many research groups are attempting to build large systems because the potential for more extensive scaling isn't there, instead they're trying to develop systems that are scalable, then we'll see a push for large systems.

It'll probably still be decades before all the hard problems are solved, but the promise of an exponentially larger computation basis for a given number of bits is too compelling to ignore.

Re:Nullo bits?

[ciderbrew]

I've not idea why this was marked down -1 informative. There is too much information

Re:Expect 20 in 20 yrs, 25 in 100 yrs, 30 in 1000

[gweihir]

I think what is difficult to ignore is the grant money you can still get for GC proposals. If the device gets exponentially more difficult to build when larger, exponentially more computing power does not mean anything. And there is rather strong indication that entangeling manipulating qbits at the same time gets exponentially harder with the number of qbits. I predict that Quantum Computers will not ever reach significant size, as conventional computers will not only simpler to build, but they do scale well for the problems Quantum Computers can solve. And there are not that many of these problems anyways. Most problems do not scale well, regardless of computing mechanism.

Re:Very cool

[GameboyRMH]

I'm against researching life extension in place of something more useful, like...almost anything else, since life extension really creates more problems than it solves (overpopulation and ethical issues vs. old people who are afraid to die (how pathetic is that)).

But apart from that, I don't care, knock yourself out.

Re:Expect 20 in 20 yrs, 25 in 100 yrs, 30 in 1000

[Interoperable]

Yes, you're right. Let's give up on investigating compelling new directions in technology because there are hard problems associated with them. That's the way forward ;-)

I'm quite certain that if the foundations for exploiting a larger computational basis are laid, the algorithms will follow.