Baby Steps Toward Quantum Computers

Mz6 writes "In a step toward making ultra-powerful computers, scientists have transferred physical characteristics between atoms by using a phenomenon called entanglement, which Einstein derided as 'spooky action at a distance' before experiments showed it was real. Such 'quantum teleportation' of characteristics had been demonstrated before between beams of light. Teleportation between atoms could someday lie at the heart of powerful quantum computers, which are probably at least a decade away from development. Researchers using lab techniques can create a weird relationship between pairs of tiny particles. After that, the fate of one particle instantly affects the other; if one particle is made to take on a certain set of properties, the other immediately takes on identical or opposite properties, no matter how far away it is and without any apparent physical connection to the first particle." Reader starannihilator adds: "Physics Web provides a good graphic summary of the phenomenon, as well as a good technical article."

Comments

Analogue vs Digital

[Nermal6693]

I think (although I'm not certain) I read somewhere that a quantum computer is like an analogue computer - where you're not restricted by 0 and 1. Is that correct?

Re:Analogue vs Digital

[Anonymous Coward]

It was just a dream, Bender. There's no such thing as two.

Re:Analogue vs Digital

[LnxAddct]

A quantum computer is completely different. The only thing in common in the binary number system. In a classical computer you have bits, either a 1 or a 0. In quantum computers you have qubits which can be a 1 or a 0 or actaully both values at the same time! This can manifest tself in amazing ways. You can try every solution to a problem instantaneously because instead of having to count throught all of the possible inputs, i.e. going from 0 to 255 with 8 classical bits, in a quantum computer 8 qubits actualy are the values of 0 through 255 all at the same time. The answer is then decomposed or observed forcing the quantum state into a final and complete solution. Some quick info for those who have no idea what qunatum anything is... an observation is essentially defined as any force that forces a quantum state to be amplified into a definitive state. Quantum entanglement occurs when two paritcles intereact for a short period of time (i.e. two photons crossing) and then go off on their own, they can travel to oppisite sides of the universe and whatever happens to one, instantaneously happens to the other. Literally, no moment of time occurs between the change, its quite amazing. If you polarize one photon, the other will instantly be affected. Also if particles A & B are entangled and C & D are entangled then if B entangles with D then A automatically becomes entangled with C. This allows for some truly amazing things. One final note, although quantum entanglement was first observed with laser light(photons), it has since been reproduced with much larger particles including ruby atoms and even bucky balls (google it if you dont know what one is)
Regards,
Steve

Re:Analogue vs Digital

[jfern]

With n classical bits, they can be of 2^n possible states.
With n quantum qubits, they can be any normalized (overall phase doesn't matter) complex vector in 2^n dimensions.
However, when you measure them, the wave-function will collapse (unless you believe in the many world's multiverse), and you'll get n classical bits.

Classical information is simply a subset of quantum information.

Re:Analogue vs Digital

[natmsincome.com]

Not so much Analogue vs Digital but rather Serial vs Parallel.

In searial you do one instruction per peice of data. In parallel you try EVERY piece of data in one instruction.

Some problems are trivial in serial but hard in a parallel and other problems are trivial in parallel but hard in serial.

Simple Example:
Iterative calculation are great in serial but aren't that good in parallel as you can calcualte the second value till you have the previous value.

The Famous example:
The big thing that quantum computers will do is make parallel problems trvial. The big two being simulations and cryptology. Cryptology is only hard because you have to try so many different combinations. Quantum would allow you to try EVERY combination at a single time. This make encryption almost useless at any key length.

It's also usefull for simulations like ray tracing and vector maths where you have a complex eqation where you just have to run for every possible variable.

So ever is a single iteration takes 1 hour for a quantum computer instead of 100th of a second for normal computers it will change the world. Breaking a key 2048 bit key will take exactly 1 hour instead of million+ years. Rendering a frame will take 1 hour on a single computer instead of 4 hours on 1000+ computers.

That being said it would be useless for Word, Excel or Firefox :-)

Imagine a quantum computer that does 5 Hz out perform a cluster that does 5 TeraHz.

Re:Analogue vs Digital

[jfern]

Quantum computers aren't quite as powerful as you make them out to be. At the end of your algorithm, you have to perform a measurement, and each qubit when measured only gives you 1 classical bit.

It's been proven that quantum computers are no better than classical computers at sorting (both O(n log n), although they are better at finding something in an unsorted database (Grover's algorithm does O(sqrt(N)), instead of O(N) classically).

No one has proven that quantum computers are faster than classical computers for factoring. We just know of a fast algorithm for a quantum computer and not a classical computer. It's likely that quantum computers are much faster there, though.

Re:Analogue vs Digital

[Medevo]

The limit of computing is, as you say, on the developer's side, no argument here. It its at least partially reasonable that when quantum computers become more available, that ingenious developers will find ways to squeeze out more power.

Moreover, at the end of the day, you still extract bits from qubits. While one day in the distant future we may be able to interact computers entirely in a quantum environment, but it's a long way off.

The real potential in quantum computers is the problems of density, power, and heating, that have plagued development of faster CPU's seem to apply on a lesser scale to quantum circuits (not that they don't have there unique problems). At the same time, quantum computers could/would suffer a lot less problems with bandwidth/time delay (light/QE info transfer).

Traditional MOSFET based transistors, while powerful (look at today's advanced chips) have been around for a while; there is no harm in looking for something new and better.

Even if quantum computers provided a liner growth curve in processing power to qubits, we could expect a greater throughput in it (due to above stated factors).

Medevo

Re:Analogue vs Digital

[spacecowboy420]

OK, maybe I'll sound like a jackass, but I gotta ask anyway. It seems to me that if you can reproduce entangled particles reliably, and you have, lets say two hosts, both with one half of the set of the entangled particles. If you were to manipulate the state of one set, and that immediately affects the state of the entangled partner on the other host, wouldn't that be the effectively TRUE wireless communication. One where the rate of communication is limited only by how fast you could read and process the set of particles that are local? Wouldn't that be as secure as it gets - media to intercept? Sure, there would need to be software to interface with the states based on the input from the hosts - but if you could do this, you could control the mars rover in realtime. Is this where this is headed, or am I confused?

Re:Analogue vs Digital

[Scorillo47]

Note that entanglement is just one approach in building quantum computers, and it is not really the ONLY approach.

Generally, a quantum computer consists in several quantum systems (for example captured particles, etc). The (quantum) state of these systems varies according to a well-known equation, called the Schrodringer equation. This is a very simple equation that describes the evolution of the system (the derivative of the current vector state) in respect to the current current state & time.

The nice thing about quantum computers is that they operate with multiple simultaneous states, therefore achieving some sort of parallelism. Basically a quantum system can be considered to have a superposition of states - it has two states at once if you want. Some of these states might converge to the same state depending on the hamiltonian or on the external interactions.

The hard part is that you never know when such a computer stops its calculation since the transformation state is fully reversible and goes on ad infinitum. If you want simply to test if the computer reached the end of the calculation, you will affect the current state. Anywyay, this challenge plus many others (for example the precision of the measurement, etc) makes quantum computing very challenging.

Still, there is a theoretical possibility that you can get a high degree of parallelism in certain configuration. A classical result from Shor (you can search on Google) shows that one of the classic problems in arithmetic - integer factorization - can be done in a polynomial time on a quantum computer. This simply means that RSA encryption can be potentially broken, irrespective to the length of the key. But we are still safe - so far nobody built a working quantum computer that would carry on simple calculations like factorizing the number 15.

On the other side, entanglement is an interesting quantum fenomenon which works like this:
1) First, you have to have a way to build pairs of entangled particles. There are several ways to do this, for example by having any quantum process that generates a pair of photons.
2) Second, if you modify the vector state of one particle, the vector state of the other one will be equally affected, regardless of the distance between these two particles!

What's interesting is that entanglement guarantees instantaneous quantum state change therefore contradicting somehow the theory of special relativity. This theory says that events cannot be 100% simultaneous if they occur in different points in space - there is a timing separation based on the particular reference chosen. Practically, no standard matter interaction can be faster than the speed of light.

But there is an exception here - "collapsing the vector state". If you measure the state of a particle, its state will collapse along one of the measured dimensions (according to certain probabilities). The corresponding entangled particle will suffer a similar change, so if you measure now the state of the this second particle you will see that its vector state has already changed - and you can even perform a partial correlation between the results of the two measurements.

In conclusion, enanglement guarantees instantaneous "interaction" regardless of the distance between these paired particles (this is why Einstein called it "spooky action at a distance" - because technically it is propagated with infinite speed). Anyway, it has be proven a while back that this does NOT contradict the special theory of relativity since this is not a standard matter interaction, like gravity, etc.

Going back to computers, entanglement is an interesting approach which might enable new algorithms or new ways to build such computers. But keep in mind that we are in the stone age of quantum computing right now...

Re:Analogue vs Digital

[cicadia]

Well, you're not a jackass, but it is a bit more complicated than that. Unfortunately, there doesn't seem to be any way to actually transmit information instantaneously with entangled particles. It's true that two entangled particles will undergo the same transitions at the same time, but since you can't predict in advance or control what transition will occur, it doesn't help you send any information to a person looking at the particle at the other end.

You're right, though, that it's about as secure a communication channel as you can get. It's actually the basis for quantum cryptography -- two people share a set of entangled photons, and they can guarantee that the measurements they make on them will be identical, giving them a shared secret key that no one can intercept. They still have to communicate over regular channels to actually send any real information, though.

Re:Analogue vs Digital

[wass]

Wow, 2 quantum computation articles on /. within two days.

I mentioned this yesterday as well, but for an idea of what qubits are you can take a look at my currently unfinished Java Quantum Computation applet. As of now one can only do single-qubit operations, but eventually I hope to have a demo of quantum teleportation (teleportation of a single qubit, or spinor, that is).

This applet will give you an idea of what qubits are. Essentially they're a 'spinor' which in quantum-mechanical terms is a 2-element discrete wavefunction. In lay terms, this just means a set of two complex numbers (properly normalized). They are also displayed in a more visible representation, called the 'Bloch Sphere'.

This applet will let you take any input qubit, and operate on it with 6 different single-qubit quantum gates, and see the resulting qubit.

Look at the two qubits represented on the Bloch sphere. The yellow vector represents the qubits. The red dot indicates a classical 'zero' and the blue dot indicates classical 'one'. In classical computing any bit can only point exactly to the red or blue dots. In quantum computation a qubit can point anywhere on that sphere.

[For the mathematically curious, a qubit is 2 complex numbers, which would be 4 independent parameters. However, the sum of the modulus squared of each complex number must be unity, so that constraint leaves only 3 free parameters. Secondly, the entire qubit can be multiplied by any arbitrary phase constant (e^i*gamma) which changes the spinor but not its relative values. Hence, there are only two parameters for each qubit that really matter, so it can be expressed in 2D, mapped nicely to the sphere.]

In classical computing there are only 2 single-bit gates - Not and Buffer (actually, I never formally studied computer science, so someone please correct me if this isn't true). 'not' flips the bit, 'buffer' keeps the bit unchanged. In quantum computing there are infinitely many single-bit gates, some of the common ones are demonstrated in the applet. Basically, these gates can control how relatively 'one' or 'zero' the bit is by the superposition, as well as change the relative phase.

Anyway, I should be adding in two-qubit operations soon (like the infamous controlled-not) and hopefully get to something worthwhile.

So this applet isn't very useful for actual simulation of quantum computation yet, but it will you give an idea of what qubits are and how they can be represented.

Re:Analogue vs Digital

[Scorillo47]

Correct - there is no way to transmit pure information through photon entanglement for example. But it is possible to use this technique to verify some information transmitted in conjunction with a separate (classic) channel.

This has two consequences:
1) First, it is practically possible to use entanglement to build networks that are 100% guranteed to transmit either correct information or error.
2) Second, since measuring any particle will necessarily change it state gives an interesting conclusion: it is impossible to tamper the communication channel that transmits entangled photons. As soon as you attempted to measure what's on the channel, the verification mentioned above (i.e. the correlation between the final measurement of the two entangled particles at the two ends) will fail!

Therefore you have a bullet proof method that will prevent active/passive attacks on the entangled channel. The technique was actually employed in practice - see this link for example.

NB - this technique still doesn't prevent attacks that fully substitute one of the ends with a completely identical device so the other end still thinks it is talking to the right person. But in combination with standard cryptography techniques for the insecure channel, this techniue is almost impossible to break. A nice overview is presented here

Re:Analogue vs Digital

[ganhawk]

Let me try to explain quantum teleportation whith what I know of it.

There is no equvvalent macro phnenomenon for quantum teleportation. But let me try this example.
Quantum teleportation is something like this..

If you have a metal box that can be broken into two metal boxes. Initially there are two colored balls in the metal box. You cannot see the balls. When you break the box into two, each ball stays in one box. You can now seperate the box by a large distance. This pair of boxes is similar to entangled pair. By obeserving the first box, you can determine the color of ball in the second box.

At the quantum level, without knowing the color of the ball, it is assumed to be in a state of superposition. so Observing the first photon forces the state on the second photon.

Prime Intellect?

[Anonymous Coward]

Isn't this the correlation effect mentioned in the prime intellect story?

In the PI universe, a Beowulf cluster of these imagines YOU!

can someone qualified answer this question

[Anonymous Coward]

Just say 20 years from now I am on my quantum fandangle computer that does sub-atomic calculations, what happens when background radiation hits the processor and flips a few 1s and 0s?

i.e. will my computer crash when there is a solar flare?
will the new "heatsinks" be lead shields?
will we need to rotate the shield harmonics? (j/k)

please... inquiring minds want to know.

Re:can someone qualified answer this question

[phoenix.bam!]

I'd have to say (not that I actually know) that there would be equal danger now from a solar flare crashing your computer as there will be on a quantum computer. But what the hell do I know? You should go ask Scotty.

Re:can someone qualified answer this question

[Anonymous Coward]

Scotty's busy trying to talk to the mouse.

"HELLO COMPUTER"

Re:can someone qualified answer this question

[jfern]

The problem with quantum information is that you can't clone (copy) an arbitrary quantum state, and you can't measure an arbitrary state without destroying the quantum information.

However, there still exist quantum error correcting codes that can correct an arbitrary error. Classically, one only gets bit flip errors. In quantum computation, you have to worry about phase flip errors, for instance instead of a|0>+b|1> you have a|0>-b|1>.

The smallest quantum code that can correct an arbitrary non located (located errors are easier) error on 1 qubit requires 5 qubits. There's a 7 qubit "CSS" code that is important for fault tolerance.

For fault tolerance, you concatenate a code with itself many times, and if your errors are independent of each other, then by doing all sorts of complicated fault tolerant techniques, you can get fault tolerance. What happens is you get a fault tolerance threshold. If your rate of errors are less than that, you can do arbitrary quantum computation with O(M) qubits in O(N polylog N) time, where O(M) is the qubits required on an error free quantum computer, and O(N) is the time required on an error free quantum computer.

Re:can someone qualified answer this question

[nihilogos]

Just say 20 years from now I am on my quantum fandangle computer that does sub-atomic calculations, what happens when background radiation hits the processor and flips a few 1s and 0s?

Quantum error correction. is a sub-field of quantum computing concerned with just that, how to effectively perform a quantum computation in the presence of background radiation and other stuff which sub-atomic thingies tend to be quite sensitive to.

The likelyhood of flipping a few zeros and ones ( and other errors which can afflict quantum bits) is very high, and in reality is more a continuously decay than an instant flip.

It has been shown, however, that this continuous decay is equivalent to flip errors and phase errors (the other sort of quantum error) occuring with some probability. That probability is 1 in 10 for most of the current experiments, compared to your box in front of you which is more like 1 in 10 billion.

Fault-tolerant quantum computing is a theory field of research concerned with how good quantum computers have to be before quantum error correction can work. The best results at the moment suggest a probability of error of 1 in 1000 is good enough. The experimenters have a fair ways to go yet.

Re:can someone qualified answer this question

[mcrbids]

That probability is 1 in 10 for most of the current experiments, compared to your box in front of you which is more like 1 in 10 billion.

Would you really think even a e-machine is that error prone?

Think about it...

2.5 Ghz * 32 bits/cycle = 80,000,000,000 - that's 80 BILLION bits per second...

Of course, that's theoretical, there's buffering delays, cache, noops, etc. But, given the theory, there'd be 8 random errors every single second.

Something doesn't sound quite right, here, especially when you figure the vast majority of computer are sold with no error correction at all on the system memory ?

I think that 1 in 10 billion is probably quite a few orders of magnitude off....

Yes, fast

[Milo of Kroton]

But what cost? Only government would want new technology this fast, maybe your NSA, that around codebreaking.

Re:Yes, fast

[tachyonmkg]

Only the five richest kings of Europe will be able to afford them.

Re:Yes, fast

[plaa]

Comparing the speed of a quantum computer and classical computer is comparing apples and oranges. Quantum computers work with a totally new set of rules, which allows some applications to make use of quantum properties.

The main property that classical computers lack is that of superposition of states. One can understand this as calculating some result starting with all possible numbers at once, instead of testing each starting value as its own. (In reality it's more complicated than this, of course.)

Some applications, eg. codebreaking, number crunching and database applications could get a vast boost out of quantum computing. Other applications may not. The most probable places for quantum computers (at first) will probably be number crunching, networking applications (quantum cryptography etc) and database applications.

For a comparison, searching an unsorted database is classically an O(N) operation, but a quantum computer can do this in time O(sqrt(N)). The best known classical algorithm for factoring a number is exponential, while Shor's algorithm does it in time O((log N)^3) (allowing polynomial-time breaking of RSA).

Re:Umm...this is old news.

[beeplet]

This is the first time anyone has been able to use atoms (as opposed to photons) in quantum teleportation.

A QM foray into the private lives of Alice and Bob

[wwest4]

Alice, instantaneously transfers information about the quantum state of a particle to a receiver called Bob. The uncertainty principle means that Alice cannot know the exact state of her particle. However, another feature of quantum mechanics called "entanglement" means that she can teleport the state to Bob.


Alice: Bob, now that our qubits are entangled, I don't know if mine's spin up down.

Bob: How 'bout I observe yours for you. How about there?

Alice: Nope.

Bob: Here?

Alice: Closer to this side of the gaussian, Bobby.

Bob: How about here?

Alice: OOOOOHHH! You collapsed my wave function DeBroglie!

Bob: Your qubit is now spin up, in case you were wondering... who's DeBroglie?

Re:Not quantum computing, but

[wwest4]

Because Alice can't know the state of the information she's sending. If she does, then the superposition collapses.

It's not intuitive, but the "collapse of the wave function" metaphor fits observation.

Spooky Action at a Distance

[www.fuckingdie.com]

What happens when quantum computers, which are able to use quantum teleportation, start to exert influence directly over the matter that makes up say a Human Brain for example. Or to make matters worse the brain accidentally starts to exert control over the computer.

"We are sorry - the application you were running has crashed because you were thinking unhappy thoughts."

or

"You have 60 seconds to close and save all thoughts before your brain will be automatically restarted"

Can we say sasser-"cranial edition"

Re:How to choose?

[ajayg]

Good question. In fact, this is one of the trickier problems to solve when coming up with a QC algorithm. The trick is, to use the phenomenon of coherent interference to yield the result that you are looking for. Interference here is basically the same as wave interference. So, after our QC executes an algorithm and finds the solution to a problem for all N inputs simultaneously, we then have to interfere our output result state (which now exists as a coherent superposition of N different outcomes) in such a way as to obtain the result we are looking for. A good example you might want to look up is the Deustch-Josza algorthm, which though useless for most practical purposes (in my opinion :-)), shows how we can use intereference in a smart way to obtain the desired result.

Re:Not quantum computing, but

[jettoblack]

What you're thinking of doing is creating an entangled pair, and keeping one particle on Earth, and keepting the other on a spaceship. Then by changing the state of the Earth particle, you could affect the state of the spaceship particle. Right?

The problem is, we have no way to choose what state the particles will go into when we observe one. Its a random outcome, and you can't acheive any communication if the output is just random noise.

Furthermore, from the spaceship's viewpoint, how do you tell if your particle's state has changed due to an incoming transmission? The only way to know would be to observe it. But, we don't know if that particle had been observed by Earth yet. If it had, then we just disturbed the state that Earth had set. If it hadn't, then we just forced it (and Earth's particle) to a random state. True, the Earth's particle will now be set to the same random value, but random values are still uselss for communication.

For it to work, you'd need a second channel of information, which could transmit some kind of key to decoding the random states into data. Of course, this channel of information would have to go FTL too, so its a Catch-22...

Re:How to choose?

[jfern]

A typical quantum algorithm puts most of the wavefunction into the state(s) that you want. By applying various quantum unitary gates repeatedly one can do this. It's kind of hard to explain exactly "why". One then measures the state, and with with probability p gets a correct answer. If p> 50%, one can repeat the algorithm a bunch of times to make sure one has the right answer.

A method to break Quantum Encryption?

[SkiifGeek]

Okay, so this is probably incorrect, but it is a train of thought. With the state of quantum encryption being that if a third party observes the key in transit, it is apparent, and the key is useless, would this have a potential application to break this encryption.

Using this method, the duplicated particles could be observed, leaving the original particles in the encryption stream relatively unmolested. Yes, it would be impractical and the equipment needed would be very distinctive and difficult to hide, but it raises the possibility.

Argh!!! NOT teleport, NOT affects.

[elhedran]

Normally I am not so pedantic but the poster repeatedly misrepresented what is happening in entanglement.

4 times in the post it was said that the particles teleport or communicate, they don't.

Its more like the particles are using the same day planner to decide what to do next.

Think of it like to processes running the same code. if they have the same inputs, they will have the same outputs. It doesn't mean they communicate or teleport.

The reason it bugs me so much when people talk as if the particles interact after they have been entangled is it leads someone sooner or later to start asking why we can't use that to beat the speed of light for communication, or a dozen other things that have nothing to do with entanglement.

I love stuff about quantum computing!

[Anonymous Writer]

Too bad I can't bloody understand any of it!

This is not good...

[Barkmullz]

Having scientist using words like "spooky" and "weird" cannot be a good thing...

Quantum Window PCs?

[dcw3]

So does this mean that all the future Windows Quanta PCs will go blue screen at the same time?

I'm kidding...well, sorta.

Re:Help me understand this!

[pclminion]

If have have two boxes... A and B, which have lids on them which are shut, and if I look in box A, and either a rubber duck, or a pineapple appears, how do I know that the contents of box B have changed? I cannot open box B to look at the contents beforehand to know when they change, because that would set the state of box A.

This is confusing. You talk about things "changing" and looking in the box to see the "contents" beforehand. In the entangled state, the boxes have no "contents" to speak of, only superposed wavefunctions. By observing what is inside the box you collapse both the superposition and the entanglement.

You are asking, how can you know definitively that, before you open one of the boxes, there indeed exists an entangled superposition inside the boxes. You cannot know this. If you open a box to observe the contents, you will never observe a quantum superposition (that would be an absurdity -- it would cause your brain to enter a superposition as well. What the heck would that feel like?), you instead cause the objects to collapse to a well-defined state.

It makes no sense.

Quite right :-) But in some way, it's all connected with consciousness and observation. It seems like our consciousness is always in a well-defined state, and this "rubs off" on whatever we observe, causing any superpositions to collapse. And even if our brains did enter some kind of superposition, would we know it? Would we perceive the superposition, or would we be two superposed people, each observing what he thinks is a well-defined state?

These are questions we probably won't have answers for for a long, long time.