Purdue Builds Quantum-Computing Semiconductor

Bfaber writes: "According to EET, Purdue has created the first examples of quantum computing in a semiconductor. The story can be read here. Read the article for further links that include an audio interview."

Comments

The link is incorrect...

[snookums]

Try this one (http://www.eet.com/story/OEG20010924S0101)

Blah, blah. Lameness filter doesn't like short posts so I'll put a little padding here. Sorry to ramble, but you know how it is...

Links

[tomknight]

Well here's another karma-whoring link for y'all - it's the news article from Purdue University

http://news.uns.purdue.edu/UNS/html4ever/010917.Ch ang.quantum.html

Tom.

Encryption...

[Peridriga]

If you havn't you should read a book by Simon Singh called the "Code Book" it essentially is a history of cryptography from beginning to end (e.g. quantum cryptogrophy)....

The effects of quantum cryptography is huge... Using a quantum computer would allow you to crack huge keys (everything from PGP, RSA, DES, TwoFISH, BlowFISH, etc.... anything you can think of) because of the essential basis of quantum physics...

Simply in laymen terms you can check muliple cases of a key (i.e. check 111111 and 111112) at the same time... Not just 2 keys but, how about 2 billion keys per second... This makes any key no matter how long easily crackable...

I promise you the NSA is up early this morning banging on doors at Purdue (hey the probably funded it anyway)....

Now don't fear... Even though it makes any code breakable it also inheriently creates an unbreakable code using the same theories...

So start writing all you stuff down and locking in a safe instead of encrypting it on your hard drive.... You data really isn't safe anymore...

Re:Encryption...

[color of static]

Only public key algorithms suffer from that level of security degradation due to QC. Factoring of a number on the order of 2^n, becomes about n operations on n qubits. Symmetric ciphers (such as blowfish, DES, Twofish, RC5, AES, etc...) only have a reduction in the keyspace needed to search. So if you have a 2^n key, you will have to search 2^(n/2) keys. While there may be a way of QC reducing this further, no current theory lends it's self to this.
Of course what will it matter when there is a backdoor, and the only security is an Oracle agent smart card issued by the government?

Obsoletes planned crypto laws

[magi]

If they manage to get quantum computing working soon, and working well, we can forget these planned anti-crypto laws. Most crypto algorithms would go useless.

With quantum computers, the only way to do crypto would be transferring huge XOR mask keys physically (or possibly with quantum encryption channels). Pretty hard.

Re:Obsoletes planned crypto laws

[Dooferlad]

If they manage to get quantum computing working soon, and working well, we can forget these planned anti-crypto laws. Most crypto algorithms would go useless.

With quantum computers, the only way to do crypto would be transferring huge XOR mask keys physically (or possibly with quantum encryption channels). Pretty hard.

For those people who don't know about quantum encryption:

If you can quantum entangle two particles and move them apart, then doing something to one, has the same effect on the other. The trick is to keep them entangled for long enough, and far enough away, for this to be useful.

If you do manage to do it though, you will have a totally secure encryption channel (you can't snoop it) with no latency. Useful stuff...

-- Dooferlad

Re:Obsoletes planned crypto laws

[Omnifarious]

Two misconceptions here:

First, symmetric key encryption is still pretty good in the face of quantum computing. It isn't as good as it was. I think the difficulty factor goes down to the square root of the original difficulty factor. For a 256 bit key, that's sitll 2^128 operations to brute force it. That's pretty secure.

Second, quantum cryptography doesn't work the way you describe.

Quantum cryptography works by generating a truly random keystream using entangled particles. Since the particles are entangled, both people can get their own particle and know the state of the other person's particle. They can't alter the state of the other person's particle in any way, but they do know it.

This allows one-time pads to be securely exchanged over a distance. If someone listens in to the entangled particle stream, this irrevocably alters it, and when both sides compare a few (not all) of their shared random bits over an insecure channel, they can detect this snooping.

Quantum cryptography does NOT, I repeat, DOES NOT allow you to communicate with no latency. The speed of light applies to the particles in the entangled stream, and it applies to subsequent communications encrypted using the information in these particles. One particle of an entangled pair can only detect the collapse of the quantum wave function (i.e. when the particle is 'read') for the other particle. No other state changes can be detected by the other particle. No faster than light information exchange to see here people, move along.

Re:Obsoletes planned crypto laws

[Dooferlad]

That's presuming you have a known plaintext. That's usually not too hard to engineer, but with careful implementation, it should actually be very hard.

I agree, but there is always a chance. Of course you could enter quantum plaintext which is trial encrypted by a quantum key and then retrieve it that way :)

Some useful background on Quantum Entanglement and Quantum Communication can be found at the Centre For Quantum Communications for confused readers (like me).

-- Dooferlad

Clarification

[CaptainAlbert]

It's easy to get confused about quantum computers, because the media hype doesn't take into account the fact that you need at least two degrees (comp sci and physics) to understand it properly... guess what, I don't have these! But I do have the first, and my girlfriend has the second. :-)

Quantum cryptography itself is not an algorithm as such, but a way of using the inherent uncertainty in the polarisation of photons to ensure completely private communication. There are some labs which claim to have such a scheme working, but it's a long way from becoming feasible on a large scale.

Basically, it works on the principle that observation changes the observed event. You can ensure a secure (non-eavesdropped) channel by makeing sure that every photon has arrived correctly. If an intruder has observed your message, then the message itself has changed (at the quantum level)! I'm really not sure how it all works either, but there is plenty of published work.

The other crypto-related quantum computing thing is Shor's algorithm. For a reasonably good explanation:

http://www.doc.ic.ac.uk/~nd/surprise_97/journal/ vo l4/spb3/

In essence, factorisation of large numbers (which is an NP complete problem on conventional hardware) can be done really quickly. This threatens RSA, Diffie-Hellman etc (anything which relies on the non-factorability of products of large primes).

I expect there's a similar "quantum" attack on symmetric encryption schemes like IDEA and DES, which would just do very fast brute force searches on the key space.

Hope this clears up some misconceptions!

Decoherence

[dido]

decoherence. Quantum dots don't seem to be very promising in this respect, as the minimum time to complete some elementary operation in them is about 10^-6 sec. while the average time to decoherence is about 10^-3 sec. (from Adriano Barenco "Quantum Physics and Computers," Contemp. Phys. 37, 375 (1996). (quant-ph/9612014). Meaning you can probably do about a thousand basic operations before decoherence makes any potential answers worthless. So what if you can pack billions of these quantum dots on a single semiconductor wafer if decoherence prevents you from getting any form of useful results because decoherence destroys any superpositions you have of your entangled quantum states before you can do anything useful. More promising so far have been nuclear magnetic resonance systems (which can take as much as several hours before decoherence sets in, only trouble is making basic operations with NMR systems takes a relatively long time too) or ion traps (if only it weren't so difficult to actually create and isolate large numbers of trapped single ions!).

Maybe the Purdue group will be able to shield their quantum dots from decoherence better than previous research on such objects has done so far. But as far as I know there is no getting around this; the best anyone can do is compute everything and read out your results before decoherence sets in.

This is not such a big breakthrough, folks. Hold onto your hats. If they can show that they can do operations much more quickly than old methods of dealing with quantum dots, or they can keep decoherence at bay longer than anyone expected, that would be the big breakthrough.

Re:Purdue?

[zardor]

What would a mass production chicken farm need with a quantum computing semiconductor?

Probably to solve the chicken and egg problem.

ResearchIndex cites and clarification

[Jayson]

ResearchIndex should be in everybody's bookmarks.

In the previous post, I wasn't quite clear (shoot me, it's 5am and I've been up all night): there are a couple of different methods that I was pulling information from. In the penultimate paragraph, the final sentence was an aside referring to a method of using entanglement to transfer the keys. The rest of the post was referring to a method using polarisations and Heisenburg. Here are the two links to the papers.

First, for the transfer by polarisations. If you are at Cal, then go ask Vazirani, it looks like he has coauthored with them: http://citeseer.nj.nec.com/bennett92experimental.h tml

Then on the use of entanglement (they do not have the actual paper, bastards): http://citeseer.nj.nec.com/context/18763/0

-j

Method:performance:power:viability

[D Anderson n'Swaart]

This article is fairly well-informed and not lacking for details on the actual experiment, but while it does briefly cover certain aspects of generic quantum computing principles, it falls short of any kind of comparisons between the different techniques currently being researched (which is fine, because I didn't expect it to delve into those areas, but I'm curious nonetheless).

Being able to understand the technicals of quantum computing, at best, only moderately well, and being remarkably bad at recalling them as anything more than vague and nebulous concepts, I am in no position to even attempt to compare the alternate approaches I have read about over the past several months, but I am wondering if anyone can either answer my questions here, or point me to an article that does. I'm not looking for immense detail; I'd rather just have an answer with basic supporting facts.

What I'm wondering:

• is semi-conductor quantum computing any more viable in the long-term than whatever other vaporous methods are being investigated?
• how different is it in terms of the equipment required, and what would this mean for scalability?
• which method of quantum computing would require the least power, and could be likely to be miniaturised the best? At the moment it seems the actual computing area is very small, but the equipment required to read output is inhibitively large
• alternatively, which method is likely to yield the best results in terms of raw computing performance, or is this ultimately irrelevant since quantum computing, if we can do it effectively using whatever method, is so damn fast anyway?
• how fast, really, would a semi-decent quantum processor be, compared to a semi-decent silicon one? (This may seem like an ignorant or even trolling question, so I apologise in advance)

One thing that caught my attention is that the quantum dots they used were 180 nm across. That's 0.18 microns, which is larger than current silicon chip lithography processes, which can etch at 0.13 microns, or 130 nm. I realise we're comparing apples and oranges, and that it is superposition (and entanglement, I think) that yields the real power of quantum processors, but I always imagined that a true quantum processor would have much smaller transistor and subsequently die sizes. I know they talk about going as small as 50 nm (0.05 micron), but iirc, IBM is researching (with some success, can someone pull the article?) similarly small lithography techniques for silicon chips too.

Any informed people in the slashdot community who can address these questions? Since I am writing a science fiction novel that integrates quantum computing, and I'd like it to be as realistic as I can potentially make it with educated guessing (hahaha, I hear you smirking already), I'd appreciate any help.

Quantum Cryptography is totally different

[billstewart]

Quantum computing can be used for cryptanalysis, letting you solve problems, such as factoring, that are the core of cryptosystems like RSA and Diffie-Hellman. Quantum Cryptography is entirely different - it's a technique for sending bits securely down a fiber, using quantum techniques to tell whether someone's tried to eavesdrop on it. This is really useful if you've got a spare fiber connecting you to your recipient and you're worried about KGB eavesdroppers, but isn't too useful in the real world.

Good reference - Brassard's Bibliography