Web Quantum Computer Simulator

Heraklit writes "As reported on Heise News, the Frauenhofer Institute of Computer Architecture and Software Technology has made available the first online quantum computer simulator - it will be simulating up to 31 quantum bits, for testing new advanced quantum algorithms. Behind the scenes, it is a 32 node Athlon 3200 Myrinet Linux Cluster with 56GByte RAM! Now imagine the computing power of a few hundred qubits, if ever constructed..."

Comments

maybe they should read /. more?

[garcia]

Wow, I really hope that they didn't put those 32 processors and 56GB of DDR RAM into use for this. Sounds like they should have read this article instead. Maybe it would have been cooler and not so grainy!

wow!!!

[the_2nd_coming]

oh my god, I hope that it is only a few years after Quantom computing is broken into that normal people can buy those suckers.

can you imagine how long it would take consumer software vendors to catch up to that power? hardware vendors would have to build in planed obsolesce just to make money.

Re:wow!!!

[Paulrothrock]

Not to mention that qubit computers will be low-power and low-heat devices. One of those puppies could fit in a cell phone. Talk about ubiquitous computing, everyone could have two or three of these systems.

Re:wow!!!

[zeath]

Unfortunately, quantum computers aren't as powerful to the giddy consumer as that cluster describes. They're capable of doing repetitive, simple mathemetical tasks simultaneously on a large number of values. It's extremely complicated how that works, but I have it written in this paper (pdf) that I wrote a few years ago. The paper was focused primarily on quantum physics for the first half (also interesting, and related to the story ran a few weeks ago on the red laser and the parallel universe theory), while the second half deals with explaining how the quantum registers work. It starts in the second paragraph of page 3, though a few terms reference previous topics from the paper. It's only a few pages long and it'll explain a lot of things (some things more technical than others) that none of the articles explained. Especially pay attention to the first full paragraph on page four, which I'll quote here:

Richard Feynman was one of the first to see the potential in quantum superposition for solving such exponentially complicated problems much faster. For example, a system of 500 qubits, which is impossible to simulate with any computer today, represents a quantum superposition of as many as 2^500 states. Each of these states would be equivalent to a single list of 500 1's and 0's in a classical computer. A single quantum operation on such a system would simultaneously operate on all 2^500 states; with a single tick of the quantum computer's clock, the operation would compute not just on one machine state, as our serial computers do, but on all 2^500 machine states at once. Eventually, observing the system would cause it to reduce into a single state corresponding to a single answer, a single list of 500 1's and 0's, as measured by an axiom of quantum mechanics. A classical super computer would take approximately 10^150 separate processors to accomplish this task in the same amount of time (which is, of course, impossible).

What I can explain without too much trouble is that the cluster is merely emulating the abilities of a quantum computer. A quantum computer, conversely, would be incapable of matching the performance of, say, seti@home on all of those machines. Emulation is taxing on any system - just ask the people who are using PearPC on their brand spankin' new computers only to get sub-G3 performance out of OS X.

Re:wow!!!

[delibes]

> (which is, of course, impossible).
"If you've done six impossible things this morning, why not round it off with breakfast at Millway's, The Restaurant at the End of the Universe?"

'nuff said.

Re:wow!!!

[Too Much Noise]

Interesting paper - as long as one does not look up the references to note that it's an embarrasing copy from the second one. Including the wrong 'single-photon interference' thought experiment which, funny enough, is traced back to Deutsch[*] - remember the 'see the multiverse with a RadioShack laser and a bunch of pinholes' article a while ago?. makes one wonder ehich answer he thinks is true now - one photon multiple paths or one photon and a whole bunch (erm ... maybe just one here?) of 'shadow'/whatever multiverse photons ...

'nuff said.

[*] cf. the web page of your second reference

Re:wow!!!

[zeath]

That is amusing. I didn't even bother to look back on my sources since then. Suppose that's a lesson for me to double-check my sources before I reassert a paper. And I will be the first to agree that it is a shameless copy - it was a random research paper for an Advanced Discrete Mathematics course, certainly nothing ever intended to be a reflection of any personal research, so paraphrasing was the name of the game. It was just a convenient pre-written answer to a lot of questions people were posting.

Re:wow!!!

[the_2nd_coming]

oh yeah, now that Feynmann is dead, stern can say what he wants and not get shot down by Feymann

Qbit algorithm explained...

[Anonymous Coward]

The algorithm in psuedo-code:

int qbit[32];

for ( i = 0; i < 32; i++ )
qbit[i] = (rand() >>30) & 0x01;

Nice rack.

[Anonymous Coward]

Nice rack, seriously. Clean, uncluttered.

Re:Nice rack.

[avgjoe62]

Nice rack, seriously. Clean, uncluttered.

It is a sympton of my utter geekishness that I imagined the blonde with the nice rack after I looked at my messy, cluttered equipment rack...

Re:Nice rack.

[thempstead]

be nice though if micheal had actually thought to check the text on the page which indicates that 32 Athlon MP 2000+ CPU's are used in the rack shown which is not what the link to the image in the submission says .....

t

guess since its not to far off

[2057]

I better get studying on Quantum Computers....

Re:guess since its not to far off

[Anonymous Coward]

I already know more than you - I watched all the Quantum Leap seasons!

You cannot imagine a beowulf cluster of these

[marat]

Since power and probably complexity to program increases exponentially.

Re:You cannot imagine a beowulf cluster of these

[FrYGuY101]

You'd be able to imagine it properly if you had a beowulf cluster of quantum compu... oh, wait.

Those crazy Perl users have beaten them to it!

[James A. S. Joyce]

It's more convenient than Web interface and has no arbitrary limits...it's a quantum computing module for Perl! There's also libquantum for C users, and QCF for Matlabbers.

Re:Those crazy Perl users have beaten them to it!

[hweimer]

It's more convenient than Web interface and has no arbitrary limits...it's a quantum computing module for Perl! There's also libquantum for C users, and QCF for Matlabbers.

I don't know about QCF, but Quantum::Entanglement and libquantum take a different approach. The perl module gives a rather abstract layer without simulating the physics of a quantum computer at all. libquantum has been designed as a gate-level simulator which allows the analysis and optimization of complex quantum circuits.

Here, the simulation goes all the way down to the quantum-mechanical description of a quantum computer. This is a computationally a harder task which explains the heavy hardware. This is nothing new and has been done before (e.g. here).

Re:Those crazy Perl users have beaten them to it!

[wass]

I am working on a quantum computing simulation in Java as well, and it's been up for a few months now. So far I only have single-qubit operations, which are useful only in explaining how qubits act and how they differ from classical bits. The meat of quantum computation doesn't really kick in until you can have multiple qubits entangled.

Anyone interested can try it out here . You can take any valid input qubit, operate on it with any of six different single-qubit operators, and then see the output qubit. Qubits are represented as both complex spinors and on the Bloch sphere.

Next up is to add two-qubit operations, then work to having a controllable demo of quantum teleportation. I'd appreciate any constructive comments, if anyone would like to add some input.

Re:Obligatory

[grahamlee]

If I did, they would collapse into a single state and be useless as quantum computers.

Finally...

[bairy]

A "PC" that just scrapes Longhorn's requirements.

Tron

[Sinful_Shirts]

"Now imagine the computing power of a few hundred qubits, if ever constructed..."

Tron? :)

Re:Tron

[DataPath]

Noah constructed a supercomputer called "The Ark" that was composed of 300 qubits by 50 qubits by 30 qubits, or 450,000 cubic qubits.

Hmmm... sounds like the basis for some cheesy sci-fi - kind of like a matrix, but to protect minds from some psycho-viral plague.

Re:Tron

[MZGuy]

"Skynet begins to learn, at a geometric rate. It becomes self-aware at 2:14 a.m. eastern time, August 29. In a panic, they try to pull the plug.."

Re:Tron

[Daverd]

"Now imagine the computing power of a few hundred qubits, if ever constructed..."

Phew, for a second there I thought you were going to mention something about Beowulf clusters...

Commercial Backing?

[Yonkeltron]

Imagine the amount of rendering you could do with that thing. It'd be like having a bazillion artists painting pixels for you. How long until Pixar and co jumps on the bandwagon? This type of technology has some amazing possibilities ergo it would be no suprise if commercial support came about very quickly.

Re:Commercial Backing?

[26199]

Quantum computing will never be useful in graphics... because each qubit only ever results in a single bit of information. Even with an unthinkably powerful 1000 qubit computer, one computation is going to result in at most 1000 bits of image.

Quantum computing is useful when you have problems which are very hard even for short answers... like the travelling salesperson problem.

Re:Commercial Backing?

[DerWulf]

yeah, but don't forget, the privacy implications could be huge ... What if my boss simulated my neural network on his pocket quantum computer, finding out how little work I do.

Quantum Observations

[Nom du Keyboard]

Well, there goes my private encryption key.

The only question left is, can a Quantum Computer Simulator handle the /. effect?

Re:Quantum Observations

[shadowcabbit]

The only question left is, can a Quantum Computer Simulator handle the /. effect?

Maybe, maybe not.

Re:Quantum Observations

[WarriorPoet42]

It both can handle /. and cannot handle /. This is known as the CowboyNeal's Cat Effect.

Re:Quantum Observations

[rsadelle]

Wouldn't that be CowboyNeal's CAT-5 Effect?

Re:Quantum Observations

[IANAL(BIAILS)]

Well, we won't know until we look...

Re:fp!

[garcia]

Aww, I have a fan club! Thanks AC! This is the greatest day of my life!

Oh wait.

For the quantumly challenged amoung us

[Timesprout]

ie me, can somebody please explain in lay persons terms what simulation of quantum processes involves?

Re:For the quantumly challenged amoung us

[shadowcabbit]

strictnein writes:
the answer I got was.... I don't know. I don't even know what the fuck I did. Just pushed buttons and two minutes later it told me I was done! THE QUANTUM POWER IS AMAZING!

Profane MuthaFucka writes:
Isn't Qbit that dude that jumps all over the pile of blocks?

maxbang writes:
You changed the outcome of the loading time of the page by posting a link to it!

If anyone can explain quantum computing to you and me, pal, it's patently obvious they don't read /.

Re:For the quantumly challenged amoung us

[Anonymous Coward]

Traditional digital computing uses the basic bit that can be on/off. Quantum computing uses a qbit that can be 0, 1, or superposition of the two. Using this formalism, one can construct simulations that are "instantious" of complex systems that are modeled using probability distributions instead of traditional statistical techniques. The problem is that now the computational work has been shifted to setting up the model for the simulation. But the model will always be "instantious" (if this was quantum hardware it would actually be instantious, but since these libraries simulate quantum computing it isn't in this case).

Re:For the quantumly challenged amoung us

[scrod98]

I thought I might be helpful by looking it up.

Now my head hurts and I still don't understand it, or any practical applications.

Please someone start an amusing rant about how [insert OS here] would do this much better. Need entertainment to make head stop hurting. Or beer.

Re:For the quantumly challenged amoung us

[WarriorPoet42]

Think of a 'standard' bit. On or off. 1 or 0. Classically what determines the state of a bit, is the charge of x atoms. (X is an arbitrary number based on the size of the conductor.) So what if instead of HUGE atoms, you used quantom particles? Particles that have more than charge, but have many properties. Leading itself to a computer that computes in octal or other bases.

Besides this obvious benefit, quantom mechanics allows for some 'weird' computing. Cases where the computer knows the result of a calcuation before you ask it. Or possibly is helped by quantom computers in other realities. Definiately weird, but rediculous amount of computing power.

Re:For the quantumly challenged amoung us

[scrod98]

But what about the challenge of decoherence when you interact with the computer? (Did I read that correctly?)

Does that mean that the quantum computer will crash if you look at it the wrong way?

Heck, I have one of those, today!

Re:For the quantumly challenged amoung us

[mikael]

ie me, can somebody please explain in lay persons terms what simulation of quantum processes involves?

From the paper, the quantum computer is based on having a number of atoms isolated from the rest of the system, with each atom being able to encode 0, 1 or unknown. Somehow, they've got to convert a desired algorithm into a combination of laser pulses to set the initial state, set the interdependencies between the Q-bits (or atoms), wait for the system to find the solution and read back the final result. Probably, setting the interdependencies is going to be the hard bit.

Re:For the quantumly challenged amoung us

[WarriorPoet42]

What this refers to is the fact that quanta do not have discrete positions. They have probabilities. 50% chance of location A or 50% chance of location B (to make it simple). The issue is that until you check, it exists at BOTH A and B. But even after you check, there are problems...

One of the fundamental principles of quantum mechanics is Heisenburg's Uncertainty Priciple. It states that you cannot know both the location and the velocity (remember that in physics velocity is both speed and direction).
Explanation (basic terms):
The smaller something is the more powerful the light you need to see it. This appeals to common sense. When you are looking at things VERY small (like quanta) the 'light' you use to look at it is powerful enough to move it. So you can know where is WAS but not where it IS. Or, you can know where it WAS going but not where it IS going. This pertains to ALL the properties of quanta.

Long story short, maybe the tree makes a sound when it falls , and maybe it doesn't. But once you check, you change the results. See the Wikipedia entry on Schrödinger's cat for more info.

Re:For the quantumly challenged amoung us

[NonSequor]

Basically this stuff can't be done in polynomial time. For all quantum algorithms you start by setting a bunch of qubits into a uniform superposition of states (e.g. if you do this to 8 qubits and then measure them, you will be equally likely to get any number between 0 and 255 as your result). Then you can use these qubits as input into a function and effectively calculate the value of that function over every possible value of the input. The trouble is that you don't get 2^n different values of the function, you get a superposition of 2^n states. When you measure the output, you'll only find out one of the values of the function. So in order to get a working quantum algorithm, you have to manipulate the quantum state until you have a high probability of measuring the state you want.

Quantum computing has other complexities. Every function must output as many qubits as it has for input. It's also impossible to make a copy of a qubit without altering the original qubit. This means that in any quantum programming langauge, all funciton parameters must be passed by reference. All functions must be invertible. This can be generally accomplished by leaving the inputs unaltered and writing the output to some scratch qubits which are set to 0 beforehand.

If you want to learn more about quantum algorithms, I suggest you read up on Grover's search algorithm. It's much simpler than many quantum algorithms and it's also proven very adaptible to other situations.

Re:For the quantumly challenged amoung us

[bgs4]

In a traditional computer, a 32-bit memory location can store a 32-bit number. In a quantum computer, a 32-qubit memory location "stores" a value for each possible 32-bit number. For example, the value stored for 0 might be 0.01, the value for 1 might be 0.25, and so on. When you actually read the memory location, there is (in this example) a 1% chance that you will read a 0, and a 25% chance you will read a 1, and so on.

The above is a little bit simplified. The probability isn't stored directly. Rather, a complex number is stored and the probability is the square of the complex number.

So if you want to simulate this for a 32-qubit number, you need to store in your classical computer 2^32 complex numbers. Each operation you carry out on your 32-qubit number must be done 2^32 times on a classical computer.

Re:For the quantumly challenged amoung us

[dr. loser]

The whole coolness of quantum computation lies in the idea of superposition. The input quantum bits don't just have to be zero or one, but instead can be in a superposition of zero and one. This is powerful in two ways.

First, in principle you can prepare a superposition of all possible inputs to your program. Run the program once. You've now got a superposition of all possible outputs that can be generated from your inputs.

Second, within the program itself, performing an operation some number of times N can lead to superpositions containing ~ exp(N) terms. That is, with a linear number of operations you can generate an exponentially large number of states.

You can read more about quantum computation here or here .

Re:For the quantumly challenged amoung us

[muskr]

Really, it's just solving complex second-order differential equations of Psi(x, y, z, t). That happens to be pretty tough.

Re:For the quantumly challenged amoung us

[bgs4]

There's Grover's algorithm, which is an O(sqrt(n)) time algorithm for finding a single marked element in an unsorted database of elements, according to this site:

http://alumni.imsa.edu/~matth/quant/473/473proj/ in dex.html

Re:For the quantumly challenged amoung us

[Isbiten]

http://www.howstuffworks.com/news-item210.htm

The superior power of quantum computers is due to their ability to simultaneously exist in several different, wavelike states, called superpositions. Conventional bits of data only exist in one of two states, a 1 or 0. A qubit can exist in a superpostion that is simultaneously both 1 and 0. To handle quantum data, a computer's switches must be able to interact with one another while maintaining these superpositions, so that the qubits don't fall back into 1's or 0's. Until now, researchers have tried to hold qubits in entangled states, meaning the state of any one qubit depends on the state of all others. Using this method, the collapse of one qubit back into a 1 or 0 would result in lost data.

Re:For the quantumly challenged amoung us

[Alsee]

Jeeze, 12 replies so far and no one gave a simple understandable answer.

Doing a 31 bit quantum simulation simply requires trying every possible combination of those 31 bits:

0000000000000000000000000000000
0000000000000000000000000000001
0000000000000000000000000000010
0000000000000000000000000000011
.
.
.
1111111111111111111111111111111

There are over 2 billion bit patterns. Their simulation use 32 computers in parallel, so each computer has to try 67 million of the patterns one by one.

The neat and powerful thing about quantum computers is that a single machine can evaluate all 2 billion patterns simultaneously. A quantum computer with a mere 300 bits would be able to compute more simultaneous possibilities than there are particles in the universe.

Quantum computation means that you can almost magically test every possible solution (a near infinite number of possibilities) using the same work/time it would take to test just one posibility.

Imaging trying to find the combination to a safe. Well, you just try dialing 0 0 0 on the safe and the quantum computer will try every other combination at the same time and spit out the one that works.

-

Re:Finally...

[mothz]

But how many frames per second will you get playing Duke Nukem Forever on it?

A-ha!

[PatrickThomson]

THe longhorn specs are actually calling for a 31-qbit computer! I knew it!

no fair!

[maxbang]

You changed the outcome of the loading time of the page by posting a link to it!

Re:no fair!

[beta21]

Not only did he change the outcome, the page has collapsed!

powerful only for certain algorithms

[Anonymous Coward]

'Now imagine the computing power of a few hundred qubits, if ever constructed...'

A few hundred qubits would be very powerful at factoring numbers and other such specialized algorithms. But as far as linux and other "normal" software goes, a few hundred qubit computer won't be any better than a few hundred bit software.

Re:powerful only for certain algorithms

[gr8_phk]

So what problems currently have algorithms for solution on a quantum computer other than factoring?

Re:powerful only for certain algorithms

[Anonymous Coward]

Your statement is accurate. However, it won't satisfy the hype-hungry slashweenies.

If the Q-Bit had gone to the other processor

[Nom du Keyboard]

32 node Athlon 3200 Myrinet Linux Cluster with 56GByte RAM!

If that had been a 32 node Itantium cluster, Intel could have boasted of doubled Itantium sales for that quarter.

Re:If the Q-Bit had gone to the other processor

[muskr]

Yeah, but who wants to pay that much for hardware?!

(INTC flamebait?)

So... I did my little quantum calculation deal

[strictnein]

and the answer I got was....

I don't know. I don't even know what the fuck I did. Just pushed buttons and two minutes later it told me I was done! THE QUANTUM POWER IS AMAZING!

Re:So... I did my little quantum calculation deal

[maxbang]

your answer is both true and false, simultaneously. The problem is in the measuring. I dare youto measure it. C'mon. Do it.

Re:So... I did my little quantum calculation deal

[Paulrothrock]

Only on /. is that funny.

Re:So... I did my little quantum calculation deal

[Anonymous Coward]

I measured it and now I'm stuck in 1955. Friggin Quantum crap. Can someone get me a weather forecast for Hill Valley in 1955, preferably showing a ligtning strike?

Qbit?

[Profane MuthaFucka]

Isn't Qbit that dude that jumps all over the pile of blocks?

Re:Qbit?

[beef3k]

D'oh! That's QBert!

Oh wait... it's a joke right?

Re:Qbit?

[ShadeARG]

No, that's Q-butt. You'd jump around all the time too if you had a little offshoot poking uranus.

Re:Qbit?

[whovian]

That would be Q*bert, and you might be able to play a Java version called JD*bert online here.

Re:Qbit?

[AntiChris]

I thought Noah built an ark that was 300 qbits long, 50 qbits wide, and 30 qbits in height...
"It's the Lord, Noah"/"Riiiiiight...."-B.Cosby

On the Horizon

[WarriorPoet42]

The techs that can come for this computing power is unimaginable. Several physicits have said that it would take a quantum computer on the scall of a contemporary computer to achieve feats such as teleportation (Star Trek, eat your heart out!)
Michael Chricton (of course) has dealt with the subject quite entertainingly in the novel Timeline. Again, I say the novel.

Re:On the Horizon

[hoggoth]

> physicits have said that it would take a quantum computer on the scall of a contemporary computer to achieve feats such as teleportation

Yeah, and all physicits know that if only we had enough computing power, magic would become true.

Re:On the Horizon

[skifreak87]

Just b/c we might have the computing power to store the states of every particle in the human body, doesn't mean we know how to do it. There's also the whole how to rebuild us from that data thing b/c of laws of physics. It's also possible that superluminal teleportation (faster than light) is impossible. Been reading some interesting stuff by a prof of mine at Princeton about his theory that three information theory axioms imply quantum mechanics (and thus, that no useful info can be transferred faster than light would be a "law" of the universe). And good call on Timeline, one of my all-time favorite books (never seen the movie). Everyone should go read it.

Re:On the Horizon

[WarriorPoet42]

I absolutely agree that superluminal transportation is completely beyond what we currently believe to be within the laws of physics. 'No information can travel faster than light' Heisenburg helps us here too.

Re:On the Horizon

[Jerf]

Several physicits have said that it would take a quantum computer on the scall of a contemporary computer to achieve feats such as teleportation (Star Trek, eat your heart out!)

References please? The fundamental problem with teleportation is the impossibility of simultaneouly (or even a reasonable appoximation thereof) introducing kilograms of matter to kilograms of other specially entangled matter, and sending the still-necessary "decoding" data stream to the reciever.

By my count that's at least three impossible things. In order from least impossible to most:

1. Sending the data stream anywhere in any reasonable amount of time. (It's that last clause that's a real killer, but this might be possible to some degree.)
2. Introducing the teleportee to the entangled matter... imagine trying to touch every atom in your body simultaneously. You can't just smoosh the teleportee into the matter, it won't work. Oh, and no fair killing the teleportee to do it.
3. Creating kilograms of entangled matter in the first place, and managing to keep it from reacting to its environment so it stays entangled... for entire seconds, days, weeks, years. Uh-uh, nope, not in this universe.

I don't see how QC helps with any of these.

(It's funny how something like this brings out all the physics fanboys who, in their ignorance of physics bounded only by Star Trek, honestly think we're making some sort of progress towards teleportation and the other impossibilities. Instead, real physics just keeps stacking up the reasons why these things are impossible, and the possibilities are receding, not advancing. Why do I have the sneaking suspicion WarriorPoet42's "several physicists" are just "some other teenager I found on a web board somewhere who likes to dazzle other fanboys with his command of physics^W Star Trek buzz-words"?)

Re:On the Horizon

[Nspace13]

the idea i like about teleporation is where you incinerate the person. read the pattern of where their atoms are from the burn and transmit the data over optics, then use nano bots to reconstruct in some cool inside out building process. but then again i dont base my ideas in actual science, i just like the idea, let someone else figure out if it is possible

Re:On the Horizon

[Jerf]

Another Physics Fanboy speaks out! Hi there, Physics Fanboy!

I read your "reference" (or at least the Google cache of it), and it doesn't even contain the word "computer", so I fail to see how you've supported the claim that QC can help with teleportation. See, your (attempted) sarcastic point was actually literally true; I do know that stuff. Evidentally better than you do, since I can describe why we aren't teleporting stuff around right now. Can you? After all, we teleported a photon years ago; why haven't we done anything significantly larger? (Maybe because it's impossible? Give the idea a fair shot.)

Anyone want to take a crack at providing a reference that actually, well, refers to WarriorPoet42's claim?

Re:On the Horizon

[Welsh Dwarf]

the idea i like about teleportation is where you incinerate the person. read the pattern of where their atoms are from the burn and transmit the data over optics, then use nano bots to reconstruct in some cool inside out building process. but then again i don't base my ideas in actual science, i just like the idea, let someone else figure out if it is possible

Burn? that sounds painful, but also impracticle, imaging being reconstructed with a perfect replica of 3rd degree burns on 100% of you. Also, the nano bots can't just work with thin air (or at least I think not), and even if they could, they'd be just to large and combersome to do the job right, since in order to recreate your brain in a way that you still know who you are, they'd need to recreate every single neuron simultaneously.

On another (more trekkie optimistic) note, quantum teleportation could be used for sending the information along, but it would be hellish expensive, since the twin particles will have to be transported the old way. Also this would only give teleportation between large base stations, a far cry from 'beam me up Scotty'.

Re:On the Horizon

[proxima]

why haven't we done anything significantly larger? (Maybe because it's impossible? Give the idea a fair shot.)

The link itself says, "no one expects to be able to teleport people or other macroscopic objects in the foreseeable future, for a variety of engineering reasons, even though it would not violate any fundamental law to do so." (emphasis mine).

In my opinion, if it doesn't violate a fundamental law, it's by definition not impossible.

Re:On the Horizon

[WarriorPoet42]

First of all, while there are MANY problems quantum teleportation (you name a few of the worst), there is always something 'impossible' that remains so until someone does it. As laws of physics are refined and rewriten, we discover what is possible, not what is not.
Second, I merely meant to state that QC was a 'next step', not a 'final step'. Excuse me.
Finally, as to QC helping the idea of teleportation along, check out Dr. Krauss, author of Phsyics of Star Trek. And whatever your problems with Star Trek, the man is well credentialed.

Re:On the Horizon

[Jerf]

No physical law prevents you from flapping your arms and going to the moon. It's "merely" an engineering issue. Nevertheless, it is impossible.

Of course, in the real world, there are other alternatives that aren't impossible for that specific example. Nothing obligates the universe to be so compliant in the general case.

Besides, the article may be wrong. Having kilograms of entangled matter may not violate a fundamental law but there may still be no possible configuration of matter and energy it this universe that can accomplish it, in which case it is still impossible... and I'd suggest that is a far superior defition of "impossible", though it is still too loose. Maintaining such matter and preventing it from ever collapsing in a universe with neutrinos may not be possible. (Not to mention the other two problems I pointed out...)

(There must also be a viable path to get from here to there; for instance, even if we have the material and energy to build a Dyson Sphere (in the latter-day conception as one big land area, as seen on Star Trek and most other sci-fi appearances of the concept) there may still be no way to get from here to there. I'm not claiming this right now, just using it as an example.)

Re:On the Horizon

[MrScience]

"Any sufficiently advanced technology is indistinguishable from magic." --Arthur C. Clarke

Re:On the Horizon

[Lord_Dweomer]

"Several physicits have said that it would take a quantum computer on the scall of a contemporary computer to achieve feats such as teleportation"

Or at the very least spell checking.

Re:On the Horizon

[Lord_Dweomer]

"After all, we teleported a photon years ago; why haven't we done anything significantly larger? (Maybe because it's impossible?"

Lots of things were thought to be scientifically impossible years ago, but thank god we didn't stop researching them or we'd never have proved people wrong.

Re:On the Horizon

[Alsee]

Uh-uh, nope, not in this universe.

Of course not. That's why scientists utilized another universe in 2208.

-

Patent Fun

[RareEYE]

Aren't these the same folks that hold the MP3 encoder patents? If they are the same people I wonder when they'll patent the quantum computing algorithms?

Re:Patent Fun

[Anonymous Coward]

The "Fraunhofer Gesellschaft" is a large german organization with over 80 institutes in different areas of research. More about it at

http://www.fraunhofer.de/english/

What about practical applications such as...

[drgonzo59]

harnessing the power of quantum computing to defend their webserver against hords of curious Slashdot geeks.

Re:What about practical applications such as...

[Welsh Dwarf]

You may be joking, but database applications (and, by extension, file servers, since a fs is a kind of database) could, in theory, really benefit from this.

Imaging a server where all the data is stored on qbits, and you just read the file you need off it. The throughput would be incredible. Or mabey use this to perform an SQL query on terabytes of data in an instant (more probable I think).

Of course IANAP and this wouldn't solve bandwith problems, but hay, I can dream.

Re:What about practical applications such as...

[drgonzo59]

That would be difficult. The best search using a quantum algorithms right now is Grover wich provides only O(sqrt(n)) speedup over the list of unsorted n items. That is a theoretical limit (as of now!), so O(1) searches are not possible.
I took a couple of quantum computing courses at my university as comp. sci electives, thinking we'll learn about the magical, ultra fast algorithms. Now I am fairly pessimistic about the future of quantum computing. So far (appart from quantum cryptography) they've only got a couple of qubits, and all the work is just theoretical. And even there, there are basically 2 algorithms to learn so far, the factoring and Grover's search. What if it turns out that it is exponentially hard to build them, in other words 1000 qubits will factor huge numbers but it might be almost impossible to have a 1000 qubits not interract with the environment even if Shor's quantum correction codes are used.

Re:Finally...

[powerlinekid]

So is the new coolness factor "But... does it run longhorn?"

...simulated?

[Cyclotron_Boy]

The scaling is not 1:1, so while it takes 32 Athlon processors with 56GB of ram, the processing power of 31 qbits is not that of the 32-processor cluster. This is an emulator, so the actual 31-qbit probably isn't quite as powerful as the hardware required to accurately mathematically model it. So while the computing power of a few hundred real qbits might be impressive, the computing power required to simulate those few hundred qbits would be extremely impressive.
-F

Re:...simulated?

[sploxx]

...while the computing power of a few hundred real qbits might be impressive, the computing power required to simulate those few hundred qbits would be extremely impressive.

Well, if you count quantum computing as computing, you can easily simulate a few hundred qbits by the identity map onto the quantum computer itself ;)

Re:...simulated?

[Big_Breaker]

In fact you'd need about 2^qbits of classical computers to directly simulate an equivalent quantum computer. That is because 2^qbit states exist for quantum computer on the road to calculating the answer. With only 32 computers involved in the simulation there must be a lot of serializing going on. Keeping track of the other states must be why they have so much RAM.

I like to think of quantum computers as doing sorting rather than calculation. This is because you can give it the output to a classically irreversible and it will "sort" or "resolve" for the correct input from all of it's various multiverse incarnation.

Re:Finally...

[B3ryllium]

Sorry, if you just imagine playing Duke Nukem Forever, you collapse it into a wave function and the game never gets finished.

Or something. Look, a monkey!

Damn inconsiderate...

[JoeLinux]

They've taken out all the fun of the "imagine a beowulf cluster of these..." by putting it in the article itself...DARN YOU SLASHDOT! DARN YOU TO HECK!

Re:Damn inconsiderate...

[Anonymous Coward]

Ah but in Heck we only handle the small stuff. Slashdot is big. I'm forwarding your request to Hell, where they can deal with it properly.

Regards,
Phil, Prince of Insufficient Light

Re:Damn inconsiderate...

[Alsee]

Yeah, but at least it isn't being built in Russia...

-

Life, The Universe, & Everything

[Nom du Keyboard]

And this quantum computer simulator contained the whole answer to Life, The Universe, and Everything...

Until somebody went and looked at it.

(Or does that need 42 Q-bits?)

Re:Life, The Universe, & Everything

[Zak3056]

And this quantum computer simulator contained the whole answer to Life, The Universe, and Everything...

Until somebody went and looked at it.

You're modded "Funny" but I actually found the post interesting. And here's why:

There's a bit on THHGTTG that goes

There is a theory which states that if ever anyone discovers exactly what the universe is for and why it is here, it will instantly disappear and be replaced by something even more bizarre and inexplicable. There is another theory which states that this has already happened.

I can't help but wonder exactly what Adams knew about quantum physics... :)

Re:Life, The Universe, & Everything

[Nom du Keyboard]

You're modded "Funny" but I actually found the post interesting.

I'm glad someone saw beyond the initial funny, to the deeper thought beneath it.

Re:Life, The Universe, & Everything

[Elder Entropist]

Something else that blew my mind with Douglas Adams' work and quantum:

In one of the books (Life the Universe, and Everything?) he explains about how the infinite improbability generator works, and he states that an artifical brain (Bambleweeney Vector Plotter?) is connected to a really hot cup of tea.

One of the problems with quantum computers is decoherence - isolating the qubits from the environment. I was reading an article where they were discussing a strategy for this by isolating the qubits in a fluid that had a strong random component to it, but where the many interactions averaged out to zero. A fluid with lots of brownian motion - in other words, a hot cup of tea would do.

You knew it was coming...

[eviljolly]

Imagine a beowulf cluster of....oh nevermind I don't feel like getting redundancy points.

Errors

[Anonymous Coward]

1) This is C code, not pseudocode.

1) There are 31 qbits, not 32.

3) Why the right shift by 30 bits on the rand()? You're AND-ing this with 0x01 anyway, so the final outcome will either be 0 or 1. Quite perplexing.

Re:Errors

[Anonymous Coward]

3) You should always use high order bits from a RNG.

Re:Errors

[aardvarkjoe]

If it's evenly distributed, why?

Because usually random numbers obtained from rand() aren't evenly distributed. It's also part of the reason why using (rand() % n) to get a number from 0 to n-1 is wrong.

Re:Errors

[jdray]

1) This is C code, not pseudocode.
1) There are 31 qbits, not 32.
3) Why the right shift by 30 bits on the rand()? You're AND-ing this with 0x01 anyway, so the final outcome will either be 0 or 1. Quite perplexing.

1) He said it was pseudocode.
2) Where's number two? (I fear all the answers to this one)

Now all they have to do...

[beef3k]

... is develop a quantum algorithm that can handle a decent amount of slashdoters!

a real QC would be 2^31 times faster

[menscher]

I know everyone's excited about this, but keep in mind that it's 2^31 times slower than the thing it's trying to simulate. That's because it can't really take advantage of the exponential speedup from working with entangled states. Or, more accurately, it gets an exponential speedup at the cost of an exponential slowdown.

Good time to fight software patents in europe.

[kunudo]

So that fraunhofer (german) cant't patent the algoritms used for quantum computing.

Zope

[Tobias Luetke]

Interesting that an institute like this uses Zope / Plone for their web server.

Erm, can somebody explain Quantum Computing?

[Anonymous Coward]

I've googled for it, and found articles and discussions on quantum computing no end, and seen the talk in computer magazines, but unfortunately none of the stuff has managed to even begin to explain to me how it really works. I just don't get the hang of it. (Maybe I'm just uncommonly thick... But I distinctly got the feeling that some of those editors weren't any better off...)

I would really appreciate it if somebody could just briefly unfold it here, in fairly layman terms. What kind of problems do you solve with it? (How?) How do you program a computer like that? Does the architecture have anything in common with "traditional" computers? How do you manufacture those computers? Et cetera, anything is welcome that you feel could help explain it...

I have understood that a "bit" in a QC can have any value at any given time, and that's usually where I fall off already... Thanks for any attepmts from you wiser folks! :)

Re:Erm, can somebody explain Quantum Computing?

[WhiplashII]

The simplest explanation (though incorrect, like any other simple explanation) is that you specify the answer, the steps taken to get the answer, and it tells you the inputs to the question. For example, you say: the two numbers that when multiplied together equal 42. The computer takes the qbits, multiplies them together, and throws out everything that doesn't return 42. That leaves you with the answer (or the question, however you look at it) of 6 and 7.

Re:Erm, can somebody explain Quantum Computing?

[linwoes]

Quantum Computing is not something to be understood in a blurb.

It is just too counterintuitive to what we observe to make sense. I'd recommend picking up a copy of A Shortcut Through Time by George Johnson.
It is a very good read and lets you understand just how different and powerful Quantum Computing is.

If you are still not sure...he describes how to build a Tic-Tac-Toe playing tinkertoy computer.

Re:Erm, can somebody explain Quantum Computing?

[theTerribleRobbo]

Or 6 and 8, depending on which book you read. :P

The problem with quantum computing . . .

[muskr]

For those of you who don't know: The biggest problem with quantum computing is that you can never extract all the information you compute. So you can process y=f(x) for 2^31 values of x simultaneously, but when you go to read y from the computer, you just get one solution, and what's worse, you don't even know which value of X it corresponds to!

Using Shor's factoring algorythm, however, you can extract one of the factors of a large number without knowing all the other factors. That would be useful for public key encryption. I wouldn't worry about your PGP key just yet though. 7 q-bit computers are incredibly difficult to make. The process used to make the 7-bit QC does not scale to larger numbers easily. 2048 bit computers are way beyond our technical skills.

On a side-note, I wonder if each computer simulates a q-bit (with one responsible for management). It would be the most obvious way to run the simulation, but may or may not be the fastest. There would need to be a lot of cross-communication since all the q-bits are entangled in any interesting quantum computation.

Re:The problem with quantum computing . . .

[Mr. Sketch]

7 q-bit computers are incredibly difficult to make. The process used to make the 7-bit QC does not scale to larger numbers easily.

What about simulating these higher qbit computers? After all, these people can simulate 31 qbits, but there's no way we could build an actual quantum computer that big. Maybe we can build a simulator for 1024 or 2048 qbits and run Shors algorithm on that. However, if this is possible, then I'm sure the NSA has one and all your private keys are belong to them.

Re:The problem with quantum computing . . .

[menscher]

when you go to read y from the computer, you just get one solution, and what's worse, you don't even know which value of X it corresponds to!

Not quite correct. Yes, you only get one. But you do know which value of x it corresponds to. You just can't pick your x value. It's part of the result, just like y is.

Re:The problem with quantum computing . . .

[muskr]

The computational intensity of the simulation is still bound by the laws of classical digital logic. The result is, therefore, that the simulation takes more time than an old-fashioned brute-force factoring algorythm (where digital logic speed-ups can be employed).

Re:The problem with quantum computing . . .

[psetzer]

I think that it bears repeating that anything beyond 7 qbits is really not doable at this moment, and it's questionable if it ever would be economically feasible. Basically a quantum computer gets exponentially more difficult to build the more qbits you add to it. It solves polynomial problems in linear time. This is the most momentous news since my Data Structures prof came up with a O(n^2) search algorithm for an array.

Re:The problem with quantum computing . . .

[Alsee]

Maybe we can build a simulator for 1024 or 2048 qbits

No, you are missing the exponential nature of the problem. This 31 qubit simulator is huge and expensive. A 32 qubit simulator would be twice as big and twice as expensive. A 62 qubit simulator would be 2 billion times as big and 2 billion times as expensive.

At roughly 200 qubits, the simulator would require all of the mass in the universe to build.

-

Re:The problem with quantum computing . . .

[Alsee]

That have figured out a powerful quantum error correction method. If we can build the basic building blocks then it seems we will be able to make a quantum computer pretty much as large as we like.

It's definitely a challenging problem, but things look promising.

a O(n^2) search algorithm for an array

Cool, can you post it? I've been working on this 12-dimentional array and... :)

-

Simulation vs. Real Quantum Computer

[russellamiller]

I'm confused.

How can a quantum computer be simulated by a normal computer? I'm missing something. I thought the whole point of building quantum computers was that they did work that regular computers were incapable of.

So this has to be a bad simulation. If this were a good simulation, there would be no point to building a qc. So why do we want a bad simulation?

Re:Simulation vs. Real Quantum Computer

[muskr]

The point is that QC with, effectively, 31-bits of combined CPU and RAM can do in a few instructions what this computer cluster (with several million times the memory) takes billions of instructions to complete.

Re:Simulation vs. Real Quantum Computer

[Kiriwas]

Same reason people run emulators of their favorite retro game system. They usually don't have a real one. If we had real QComputers we wouldn't need the simulator. However, we don't have very powerful ones, and the few we have have VERY limited access. So why not perfect the algorithms before the computer gets here. Computer science is no more about computers than astronomy is about telescopes.

Re:Simulation vs. Real Quantum Computer

[Anonymous Coward]

The class of problems a quantum computer can solve is exactly the same as the class of problems a normal computer can solve. The only difference is that a quantum computer may be able to solve some problems much faster.

Re:Simulation vs. Real Quantum Computer

[Thowllly]

No, a normal computer can do anything that a quantum computer can do. The whole point of quantum computers is that they can solve some problems much faster. So there might be some tasks where a normal computer will have an exponential increase in computing time as the problem size increases, while a quantum computer might do the same in linear time.

Re:Simulation vs. Real Quantum Computer

[SamSim]

It's a half-good simulation which enables us to start figuring out the total mind-job quantum algorithms that we'll be using once we get the real thing working. 's called research.

OMG! We slashdotted a QUANTUM computer! ;-)

[PaulBu]

No, really... First time in the history of mankind... ;-)

Paul B.

Re:So let me get this straight....

[Kiriwas]

Computer simulations of computers are nothing new. Just about everything is simulated before its ever made. My current research will be in simulating microarchitectures.

i, for one...

[sxtxixtxcxh]

... welcome our new simulated q-bit overlords.

Re:Finally...

[NanoGator]

"A "PC" that just scrapes Longhorn's requirements."

Giggle giggle, snort snort. We're so good at recycling other people's jokes, I wonder why none of us have girlfriends?

A question...

[polyp2000]

What I dont really get about this, is that if a quantum computer can be simulated using normal computers, then whats the big deal about quantum computers in the first place? Of course there are all the textbook theories about the benefits of a quantum computer etc. But my limited understanding of quantum physics suggests to me that

a) its impossible to simulate a quantum computer this way.
and
b), if it is possible then it cant possibly behave in the same way as a real one.

nick
(waiting for a physicist to enlighten me...)

Re:A question...

[muskr]

A 31-bit QC can accomplish in a few instructions what takes this mainframe several hours.

- C

Re:A question...

[muskr]

I suppose I should also mention that (a) it _IS_ possible to simulate a quantum computer, but it is MUCH more computationally intensive (in terms of the number of instructions). And, (b) the simulated computer behaves much like the real thing, but at a much slower pace.

It's a virtual computer

[HarveyBirdman]

It computes really fast as long as you don't actually want the answer.

Nohah

[GodWasAnAlien]

"Now imagine a few hundred qubits, if ever constructed..."

Re:Finally...

[JohnHegarty]

a sarcastic slashdot reader... you must be a killer with the ladies...

Re:Finally...

[NanoGator]

"a sarcastic slashdot reader... you must be a killer with the ladies..."

There are plenty of ladies that think of me as being like a brother to them. Now all I need is a promotion and I'll finally hit second base!

Riiiiiiiiiight! What's a qubit?

[kfg]

"Let's see, I used to know what a qubit was. Well, don't you worry about that. Just get some particles, build it."

The Wikipedia articles linked to below will certainly get you started, but they will make your head hurt.

To ease the pain in your head I recommend Nick Herbert's Quantum Reality, a popular title, but clear, concise and accurate.

There are a lot of popular works on Quantum Mechanics, but they all play the "pick any two" game with clarity, concision and accuracy. Herbert's is the only one I've found that nails all three.

One of the things that I particularly like about Herbert's book is the way he makes it explicitly clear that various models built upon interpretations of QM are a)interpretations, not QM itself and b)exclusionary.

QM presents certain logical ambiguities and paradoxes when we try to interpret it into the common world of understanding. Various models have been made to to try to deal those issues. Popular "philosophers" like to mix and match these interpretational models, believing they're a)all really the same interpretation and b)Quantum physics.

"So there I was, cruising along faster than light, backwards in time through the multiverse. . . "

But you can't do that, take one from column A and two from column B. Each interpretation is a logical structure unto itself and if you accept the multiverse interpretation adding elements from some other interpretations actually breaks the model's relation to QM.

The above 'quote' is like saying:

"So, I calculated my trajectory by Newton's Laws, but banged into a crystal sphere of Mars because I neglected one of the epicycles and didn't correct for General Relativistic forces. There's a chance I misread the initial conditions data from the chicken entrails as well."

Anyway, just read the book. It'll make you a better person, or at least a person with a more accurate view of QM than nonphysicits who haven't. Just 250 pages, so it's not even some huge tome that takes a multimonth commitment. Like I said, it's concise. Like a good O'Reilly book.

KFG

Re:Riiiiiiiiiight! What's a qubit?

[kfg]

As far as quantum physics is concerned, we might expect that it behoves those who seek quantum truths to look at each interpretation and build a common framework that is able to accommodate the essential elements of each.

Obviously a common, underlying framework does exist. So, what is it?

Quantum Physics.

KFG

Don't waste your time

[Colonel Panic]

It'll be years before actual quantum computers exist. Right now they implement them using NMR machines and they're only able to realize a handfull of qbits. So unless you want your desktop machine to double as an MRI machine quantum computing just isn't going to be cost effective for decades (if ever).

Subtle effects of the QC

[mveloso]

One unintended side effect of the QC has been that answers started to show up before questions were put in. Researchers are investigating, but suspect they already know the answer...

Re:So let me get this straight....

[hopemafia]

Yes, I know everything gets simulated these days...I was just being somewhat snarky....

So this simulation allows people to run quantum computing algorithms on a non-quantum computer...basically it's a quantum computer emulator?

Re:Finally...

[aardvarkjoe]

Now all I need is a promotion and I'll finally hit second base!

Is that the one where she gives you the root password to her home computer?

Re:My first impression when reading this was an im

[sxtxixtxcxh]

and still waiting for duke nukem forever. ;)

Re:Finally...

[Anonymous Coward]

Not to burst your bubble, but being thought of as a brother is a long fucking road to sexual contact, unless you're in West Virginia.

Quantum Fantasy

[Colonel Panic]

Having just finished a class in Quantum computing I have these observations:

1) Right now most of these quantum 'circuits' are implemented on NMR machines. They can realize a handfull of qubits. Not very cost effective. Unless you want your computer to double as an MRI machine (hey, you could rent it out every night!) it's not going to cost effective any time soon.

2) Quantum Cellular Automata (QCA) - not strictly quantum computing, but a very interesting and potentially realizable (as in they might actually be able to fabricate these in the next 10 years or so) computing paradigm. The big advantages over current logic families (like CMOS): there is no current flow hence the power dissipation could be miniscule. They switch at Terahertz rates. QCA circuits are very small ( a majority gate in less space than a current CMOS transistor).

3) Put the word 'Quantum' in front of something and it suddenly has a certain cachet.

For the time being, most of this stuff is fantasy. At most we can build actual quantum circuits (not simulated) which have maybe 10 gates or so which isn't too useful and the implementation technology is extremely expensive (not to mention large and power hungry). QCAs may actually lead to something real - but they're not really quantum gates.

Quantum cluster naming convention!

[isotope23]

I hereby recommend when Quantum clusters come out we call them :

Grendel Clusters!

Obscure joke I know but hey, I'm a geek....

I just used Quantum Computing to get a first post!

[teamhasnoi]

Unfortunately, I changed the outcome by measuring it, and am left with a -1 Insightful rant about 'Lawn Dart Deaths: Fact or Fiction?'

Damn you Quantum Computing! Your seemingly random results have cost me everything!

Random numbers

[Kallahan]

Say we do take a qbit and we just look at it from time to time, since it can be any number at any given time, can this be used to generate truly random numbers, just set up a qc and check qbits from a function's output that allows for an equal probability of getting the right state?

640 Qubits is enough for anybody.

[Anonymous Coward]

/obligatory quote for future generations to laugh at.

Re:Finally...

[Anonymous Coward]

Is that the one where she gives you the root password to her home computer?

No, that's where you finger her box, grep a couple times, and then you stream into a secure tunnel. And if you're feeling particularly nasty, you can look for a trojan and go through the back door.

Yeah he did...

[isotope23]

If only we had STARTED with Grendel clusters.... ;-)

I suppose we could call them Gilgamesh Clusters instead.

Answer to life, the universe, everything

[EssTiDee]

All the processing power in the world doesn't change the answer: 101010

But maybe a system like that could spit out 42 a little quicker than my old dual celeron.

Quantum Architecture: A Survey for Non-Physicists

[antistatickid]

I wrote a paper which might be helpful, which is geared specifically towards compsci majors (that's me) without a physics background. It's posted here, along with slides:

http://www.node99.org/projects/qa/

It's geared towards the architecture aspects of quantum computers, with a gentle intro to the physics behind it.

If this is quantum computing then...

[abertoll]

... what do you call computer circuits which are constructed from quantum dots? I thought that's what "quantum computing" was but I guess I'm wrong. Is that just really really tiny regular computing?

Google Translation of Heise article

[Elvon Livengood]

Simulated quantum computer in the InterNet

Fraunhofer Institut for computer architecture and software technology ( ROOFRIDGE ) placed a quantum computer simulator on-line accessible by Webbrowser . The simulated machine can with up to 31 Qubits so mentioned work and is help to develop new algorithms and circuits for quantum computers.

Technical details of the hard and software describe the scientists in a detailed essay on the Website. Behind the simulation by Myrinet a coupled Linux cluster with altogether 56 GByte puts main memories.

Quantum computers are able to solve computing problems very fast at those conventional computers the teeth break off themselves -- for example the factorizing of very large numbers. They can do that, because they work with Qubits so mentioned instead of with bits. A Qubit takes both binary conditions at the same time; an arithmetic operation at a register from Qubits affects therefore all values at the same time. Each selection of the result destroys however the simultaneousness (or superposition) and reduces it to only one value.

Therefore hardware is, which can manipulate the sensitive Qubits, it however on the other hand as well as possible before the destructive external world influences protects for material quantum computers necessarily on the one hand. On the other hand completely new algorithms are necessary, with which the final result contains to a certain extent all solutions. One of it is the factorizing algorithm of Shor .

Not quite.

[warrax_666]

You should always use high order bits from a RNG.

It depends entirely on the particular RNG you are using.

From man 3 rand:

The versions of rand() and srand() in the Linux C Library use the same random number generator as random() and srandom(), so the lower-order bits should be as random as the higher-order bits. However, on older rand() implementations, and on current implementations on different systems, the lower-order bits are much less random than the higher-order bits. Do not use this function in applications intended to be portable when good randomness is needed.

(emphasis mine)

So:

1. If you are on Linux, use rand() -- no need to shift, low bits are just as random as high bits.
   
2. If you are not on Linux, find a usable implementation of a good rand() and don't use the C library one (they are usually broken in various ways).
   

Re:Finally...

[Lord_Dweomer]

You could never figure it out because the instant you tried to measure it the computer would reboot.

Oh wait, that's cuz its running Windows Quantum Edition.