First Quantum Byte Created

gila_monster writes "Juice Enews Daily is reporting that the Institute of Quantum Optics and Quantum Information at the University of Innsbruck in Austria has created an entanglement of eight quantum particles, yielding a quantum byte or 'qubyte,' or eight qubits. The formal paper was published in the December 1 issue of Nature. A qubyte with eight ions provides a computing matrix of 65536 mostly independent elements. No word in the article about whether they were able to actually use the qubyte for computing."

Comments

no word in the article

[Anonymous Coward]

No word in the article about whether they were able to actually use the qubyte for computing

I think we can be sure that if somebody had unlocked the secret of quantum computing there's a chance they'd say so at some point.

Re:no word in the article

[David Hume]

I think we can be sure that if somebody had unlocked the secret of quantum computing there's a chance they'd say so at some point.

Unless that someone worked for the National Security Agency.

Re:no word in the article

[bhiestand]

Did you ever see "Mercury Rising"?

Yes, but I thought it was rather unrealistic. Any self-respecting cryptologic organization would give the young man a security clearance and let him work his magic. If they really think he's a threat, their best bet would be to lock him in one of those windowless buildings and make him check his email 500 times a day. That'll destroy him mentally faster than a lobotomy.

Re:no word in the article

[k1773re7f]

Yes. That is specifically why I have become confortably numb.

Oh good, more E-Mail

Re:no word in the article

[mikael]

Have a look at the tutorials at QuBit.org. The general principle is that the QuBit computer uses constructive interference between the qubits to generate a new state that is closer towards a solution, and eventually reaches a final state. This article describes how to implement Quantum Cryptoanalysis using a Quantum Fourier Transform.

As an example, imaging the qubits were discs of polarizing glass being rapidly spun by electric motors. You could test the state of each bit by having a set of lasers on one side to a beam of light through the discs to a bunch of light sensors on the other side. Depending on the states detected by the light sensors, the motors are used to adjust the rotation or position the discs. If you get the right feedback system, you will iterate towards whatever solution needed.

The only difference is that the quantum computer would be instantanous.

Re:no word in the article

[Mjec]

I think we can be sure that if somebody had unlocked the secret of quantum computing there's a chance they'd say so at some point.

Ummm... not quite. There's lots of quantum computing currently being done - 4 qbit computers exist in several places (or can be brought into existance on demand, anyway). Quantum computation requires entanglement and manipulation of entangled bits. Well, the former is the hard part - that's what's been managed here. A major step forward - I recall 6 qbits was the record about a 18 months ago. Entangled bits are quite delicate - so that's the next challange. Now that they can entangle this many bits, they just need to manipulate them. That'll come with time.

Quantum bytes still decryptable?

[LiquidCoooled]

Wasn't there some news recently that the so called quantum bits could be read without disturbing their state.
Which would either break quantum theory, or would mean they are just fabricated bits of information and not quantum bits at all.

The article was here

Re:Quantum bytes still decryptable?

[Silverlancer]

Read the post here. It (and a few responses to it) describe why this doesn't violate quantum theory.

Re:Quantum bytes still decryptable?

[LiquidCoooled]

Doesn't Quantum theory say you cannot read the state without disturbing the state?
The act of finding the state of a quantum bit collapses the quantum wave and obtains a result, ie you can find out what the value is now, but that may disturb what the value was going to be leading to possibly incorrect answers.

Qubits as described by modern phsyical science do not sound like true theoretical quantum bits and just sound more like tiny transistors.

obligatory bill cosby quote

[baldass_newbie]

"God, what's a qubit?"

Que?

[Rhinobird]

A qubyte with eight ions provides a computing matrix of 65536 mostly independent elements.

Wouldn't a qubyte just provide an indeteminate number of somewhere between 0 and 255 zombie cats?

Seriously, how do they get a 16 bit number out of an 8 bit qubyte?

Re:Que?

[L0phtpDK]

Umm... No.

One qubit has four states. So its actualy an 8-qubit integer.

(go through the powers of x^4: 4,16,64,256,1024,4096,16384,65536)

Re:Que?

[marol]

Don't you mean 4^x?

Re:Que?

[Ruberik]

A qubit has an uncountably infinite number of states: choose any two complex numbers A and B such that |A|^2 + |B|^2 = 1, and they define an allowed qubit. On the other hand, when you measure a qubit's state, you can get one of two results: 0 (with probability |A|^2) or 1 (with probability |B|^2).

I can't find the original article, so I don't know where this 2^16 business is coming from, but I assure you that a qubit does not have four states -- the only useful numbers for counting a qubit's number of states are infinity (quantum states) and two (possible measurement results).

If someone can link the paper this comes from, I'd be interested in reading it: I'm doing a MSc in quantum computing right now, so I might be able to decipher the source of this 2^16 stuff.

Re:Que?

[Tony Hoyle]

That doesn't make sense at all.

If a qubit is both 0 and 1 at the same time it allows for precisely 1 state (which is either 'not useful' or 'completely random' depending on your point of view.

To store data you need at least 2 independent states. That still leaves you the problem than you can't store 65536 values in 8 bits.

Re:Que?

[maxwell demon]

If I am understanding this wrong, please correct me :).

You understand this wrong.

A qubit indeed can have one of a continuum of states. For example, if you think of the photon polarisazion, each linear polarization direction corresponds to a distingt state, and then there are the circular and elliptic polarized states as well. Indeed, you can map the states of a qubit onto a sphere (embedded in ordinary 3D space), which is called Bloch sphere. Every point of that sphere corresponds to a (pure) state of the qubit. (Note that the Bloch sphere is not the Hilbert space, but for single qubits, it's IMHO much easier to understand things in the Bloch sphere picture)

Now if you measure, you basically choose a direction on that spere, and you get just one of two results. e.g. if you think of the sphere as Earth's surface, and let's assume you have chosen the direction of the Earth's rotation axis for measurement, then if the state of the qubit (before measurement) is actually the North Pole, you get with certainty one result (which, for obvious reasons, I'll call "North"), and if the state is the South Pole, you get with certainty another result (which I'll now call "South"). However, even if the state is something else, your measurement will never give anything but "North" or "South". The probability to get "North" grows the closer the state is to the North Pole, and equivalently for the South Pole. If the state is at the equator, the probability of getting North or South is the same, i.e. the result of your measurement is completely unpredictable.

Now the funny thing is that after you measured North or South, for an ideal quantum measurement, the state actually is the corresponding Pole, no matter what it was before.

If you map the states described by the article with the Bloch sphere, and say you map the states 0 and 1 to the North and South pole, then the states you named `0 and `1 would be two antipodal states on the equator, say on the zero meridian and on the 180 degree meridian (unlike in the hilbert space, the directions now are not in 45 degrees, but actually orthogonal). That is, if the state is `0 or `1, then any measurement in the north-south direction will give completely unpredictable results. Of course if you choose the direction of the `0 and `1 states (I'll call that the equatorial direction from now on), then those states will create a predictable result, while the North and South pole states will get completely unpredictable results.

Now the nice thing for encryption is that if you don't know if the state was prepared in the North-South direction or the equatorial direction, there's no way for you to know if what you got for a measurement is a prepared state, or just random garbage. Moreover, since measuring in the wrong direction changes the original state (and therefore destroys the information which was originally in there), you'll be able to notice if someone tries to eavesdrop your connection.

Re:Que?

[Pollardito]

very, very enlightening. not that i understood at all what you said, but now i understand perfectly how my dad feels when i explain the PC on his desk

Re:Que?

[milimetric]

I've read the posts here and I can point you to one source that I know is accurate, easy to understand and in my opinion beautiful:

N. David Mermin

This man is a genius. He can also explain his genius which makes him quite unique. I took a class of his and actually understood some stuff. His basic goal is to explain quantum computation to CS students. More on topic, here's the skinny on qubits:

Chapter 1 of his intro class

I really wouldn't do justice to the ease with which he explains things to attempt to summarize, but hey, what's slashdot for:

Basically skip ahead to part C if you want to jump right into it. It helps if you think of Classical bits as vectors in a two dimensional space. (0,1) and (1,0) would represent 0 and 1 as we normally think about them. So then think about Quantum bits (qubits) as (a,b) which is just a superposition of the two classical bits with amplitudes a and b which are complex scalars. The only condition is that the qubit is a unit vector in two dimensional complex vector space, or in short |a|^2 + |b|^2 = 1. Now more to the point of this thread, if you go to section 1..62 you can see that n qubits make up a computational basis (or classical basis). So, the answer is, there's not really anything like simple 0,1 states for qubits. The truth is more complicated but once you start looking at how to take advantage of qubits, a lot more beautiful in my oppinion.

Re:Que?

[tendays]

I was also wondering where this idea that a qubit has four states was coming from. It would be like saying there are four directions on a plane (north west south east) while of course there are an infinite number of them.

Actually I think the confusion comes from the fact that quantum cryptography (key exchange) is only using those four different states mentionned by L0phtpDK (btw that looks more like a password than a username are you sure you didn't swap them? :-)

Mostly independent?

[Rob Kaper]

The phrase "mostly independent" doesn't sound completely reliable to me in a world where a single 0 or 1 can change the entire meaning of data or functionality of software.

Still, with some engineering experience it's easy to fill in what the article omits. Science moved forward and technology implementations will catch up and find a way to overcome issues like these. In fact, some data mirroring with checksums might already be more than sufficient and quantum particles offer sufficient improvements in data/space ratios that duplication should not be a concern.

Re:Mostly independent?

[spot35]

We are qubit 7 of 8 or computing matrix 65536. We will will add your cryptographic and entangled distinctivenes to our own. We are the qubyte. Resistance is uncertain.

Getting there...

[meringuoid]

... Eight qubits? ISTR that Shor's original quantum error correction code requires nine, and there are simpler codes requiring fewer. We're getting here into a scale where some very interesting features of quantum computation can be demonstrated.

Why eight?

[pubjames]

Why did they choose eight 'bits' for their quantum 'byte'? For historical reasons, or is there a logical reason to choose eight? Why not seven, or 42?

I'm not being entirely frivolous - I understand quantum computing is radically different from today's architectures and so don't understand why they are choosing a byte size based on what seems to me to be historical factors.

Re:Why eight?

[grimJester]

Likely they've tried to get as many bits as possible and just now reached eight. Since eight bits are a byte, eight bits are a newsworthy milestone.

Re:Why eight?

[glwtta]

Why did they choose eight 'bits' for their quantum 'byte'?

They probably felt that 7 wasn't enough and 9 was too many.

Re:Why eight?

[City Jim 3000]

The extranous bit(s) are evil of course.

Re:Why eight?

[akaina]

AFAICT, a byte denotes 8 identifiable positions (not to be confused with states). Each position has traditionally had 2 possible states. If quantum theory allows 4 states per position a qubyte can have 65536 permutation states.

Re:Why eight?

[khallow]

Powers of two will continue to be significant. It still shows up in algorithms (eg, the quantum fast Fourier transform).

Re:Why eight?

[cev]

The work presented in the Nature article represents an incremental step towards applied quantum computing. There is no mention of "byte" in the Nature article. I suspect that the use of "byte" in the linked article is an abstraction created by a semi-technical promotional writer.

The primary interest of the result is demonstration of the fidelity of 6, 7, and 8 particle entanglements. No applied computing is done, nor is there any particular reason why they stopped at eight particles except that it appears to be the practical limit of the current engineering.

CV

Re:*Ominous thunder*

[meringuoid]

Unless quantum cryptography gets there first. The race is on.

Quantum cryptography already did get there first.

A few more..

[Renraku]

We need a few more before quantum porn.

Think about it..any kind of porn in one file..

Re:A few more..

[aug24]

Mmmm, quantum porn. Super-position, entanglement and some guy with a pussy.

J.

Re:A few more..

[syle]

But is the pussy alive or dead? Beastality or necrophilia are two completely different areas of porn and discriminating viewers need to know!

Think of the cats!

[Anonymous Coward]

Let us all take a minute to reflect on all the cats who died in support of this research.

Or maybe they didn't.

Star Trek School of Programming

[UncleAlias]

"That's not a bug, that's a quantum singularity!"

New frontiers in computing

[Urusai]

Maybe we can finally figure out what happened to that dang cat.

Re:New frontiers in computing

[indifferent children]

New physics textbook: 101 Uses for a Dead or Not-Dead Cat

But...

[cshank4]

But will it run Linux?

Re:But...

[maxwell demon]

It will be in a superposition of running Linux and not running Linux, until you look if it does.

And God Said to Moses...

[craznar]

... build a Linux Box 40 Qubits in size....

Why, Oh, Why?

[tcdk]

Do we really need this? I can't imagine how anybody will have usage for more that four qubits anyway. When will the madness stop?

Re:Whats a Qbit?

[centie]

A qubit is a superposition of two states, a 1 and a 0 if you like. So it containes some 0 and some 1, or written as a|0> + b|1>, where a and b describe "how much" (more accuratly the probability) of 0 and 1 in the state. a and b are in general complex numbers. One qubit has then 2d hilbert space, 2 quibits 4d and 3 quibts 8d etc. So 8 qubits has a 256 dimensional space for its complex amplitudes (a and b etc) to inhabit.

Quantum Calculations

[mustafap]

My laymans understanding of quantum computing is that it will enable massively parallel calculations to occur simulataneously.

The problem however is that you get all the answers simultaneously, and that the *real* problem is then finding efficient algorithms to search the results space.

Could someone who actually knows what that all means dumb it down to our level, and explain how quantum computing will actually be useful?

Re:Quantum Calculations

[centie]

You've kind of answered your own question..

The massive parrallel computation with a single element means you can solve *certain* problems in, for example, 2n instead of 2^n steps. But yes, then you get a bit matrix of answers, and reading them all out takes the same amount of steps as classical computing. But, your only usually intristed in some of the answers, so you can then use another algorithm (eg Deutsch-Jozsa) to read those out, again faster than classically.

So you get a substantial decrease (ofton exponential) in the time taken to solve *cetain* problems. Some of these problems would simply be impossible to solve in any reasonable timescales (eg milennia) using classical algorithms.

Here's a no-b.s. article on quantum computers

[putko]

I found this at Caltech, a piece on quantum computers. I've never really taken quantum computation seriously -- it just seemed too far-fetched. If they've really got 8-bits, maybe quantum computing will matter in my lifetime.

From reading the piece, it sounds like we will have some major problems with our current cryptographic systems if quantum computers become available.

Make no bones about it, Calcium works

[digitaldc]

"With a trap using magnetic fields they captured eight calcium ions, lined them up, and set up them in "W states" using a complicated laser technology"

Calcium again coming to the rescue to provide structure for a complex system. What would people or quantum computers be without it?

Re:Make no bones about it, Calcium works

[NatasRevol]

What would people or quantum computers be without it?

Flexible?

yay!

[3-State Bit]

I was born in 1983, but now I can re-experience even advances in computing that happened in the seventies and before! Cabinet-sized hard-drives that hold a couple of megabytes? Quantum computing is at A FEW QUBITS! I doubt many people here lived through the ENIAC (and realized what it meant at the time), but that's exactly what my grandchildren will be hearing from me. Granson, back in my day we had EIGHT QUBITS! Not qubytes, QUBITS, sonny boy, eight of 'em. Like this: one, two, three, four, five, six, seven, eight. Total. And that was state of the art. It was a research demonstration! And we liked it!

"There is a world market for 4, maybe 5 quantum computers."

"512 kiloqubytes outta' be enough for anybody!"

Etc, etc, etc. WHOOOHOO!!! I was there at ground zero, baby!!! In ought six!!!!

What do you mean ought-six, grandpa? "I mean 2006, granson".

"Whoa! When were you born?"

"I was born in the LAST MILENNIUM, GRANSON"

"Did they have cars?"

"Just road ones."

"What about Google?"

"yeah, but it wasn't like today. Man I wish I'd have held on to that stock tho'..."

No word?

[Syberghost]

Actually, there was an announcement, but they used their qubit to crack your ssh key in five seconds and deleted it from your email.

Re:Oblig

[maxwell demon]

1) Are they certain?

I guess they produced an eigenstate of the atom number operator, therefore they should be quite certain.

What do qbit bytes taste like?

That of course depends on what they are made of :-) Now it's very likely that their qubits only contain the traditional flavours up and down, because particles with strange flavour tend not to be very stabile.

So is this cat dead or what?

I just looked: It is dead. However, now I have problems with PETA activists from Copenhagen who claim I killed the cat by looking ...

Re:Is it just me

[coolGuyZak]

Your perspective of 2900 AD/CE is obviously flawed. In 2900, there will be no tomatoes or anchovies due to global warming. Ketchup won't exist.

And that is completely ignoring the inevitable triumph of ID...

So,

[2names]

When do we get an Improbability Drive?

Scalability

[Darius Jedburgh]

People have been expecting quantum computing to take off in a big way but after a couple of decades of research we still have only machines with a handful of qubits. I claimed from day one that the difficulty of building a quantum computer with memory N goes up exponentially. Because of Moore's law type effects our ability to build computers goes up exponentially. The net result is that I expect the memory of quantum computers to go up linearly over time, not exponentially like classical computers. I think we're seeing this borne out over the years. So don't expect quantum algorithms to crack codes any time soon. For what it's worth, I think the claims of scalability in the article are BS - but we'll see...

Let's apply Moore's Law inappropriately!

[abb3w]

The initial report of IBM deploying a 7-qbit quantum computer came out December 19, 2001. The 8-qbit result from TFA was first reported (from a Google News search) November 30, 2005-- roughly four years. This gives a doubling period of roughly 20 years (7485 days).

Which means there should be a 16 qbit machine by 2025, the 32 qbit machine by 2045... hmm. How unhelpful.