backtop


Print 77 comment(s) - last by Visual.. on Feb 19 at 5:48 AM


A view of a 16-qubit processor mounted in its sample holder

A picture of the Orion chip’s sample holder attached to a Leiden Cryogenics dilution fridge

An optical picture of the Orion processor with 16-qubits
Canadian company D-Wave shows off technology that promises to give quantum computing capabilities to mainstream industry

Canadian firm D-Wave Systems unveiled and demonstrated today what it calls “the world's first commercially viable quantum computer.” Company officials announced the technology at the Computer History Museum in Mountain View, California in a demonstration intended to show how the machine can run commercial applications and is better suited to the types of problems that have stymied conventional (digital) computers.

The demonstration of the technology was held at the Computer History Museum, but the actual hardware remained in Burnaby, BC where it was being chilled down to 5 millikelvin, or minus 273.145 degrees Celsius (colder than interstellar space), with liquid helium.

Quantum computers rely on quantum mechanics, the rules that underlie the behavior of all matter and energy, to accelerate computation. It has been known for some time that once some simple features of quantum mechanics are harnessed, machines will be built capable of outperforming any conceivable conventional supercomputer. But D-Wave explains that its new device is intended as a complement to conventional computers, to augment existing machines and their market, not to replace them.

To make the technology commercially applicable, D-Wave used the processes and infrastructure associated with the semiconductor industry. The D-Wave computer, dubbed Orion, is based on a silicon chip containing 16 quantum bits, or “qubits,” which are capable of retaining both binary values of zero and one. The qubits mimic each others’ values allowing for an amplification of their computational power. D-Wave says that its system is scalable by adding multiples of qubits. The company expects to have 32-qubit systems by the end of this year, and as many as 1024-qubit systems by the end of 2008.

"D-Wave's breakthrough in quantum technology represents a substantial step forward in solving commercial and scientific problems which, until now, were considered intractable. Digital technology stands to reap the benefits of enhanced performance and broader application," said Herb Martin, chief executive officer.

Quantum-computer technology can solve what is known as "NP-complete" problems. These are the problems where the sheer volume of complex data and variables prevent digital computers from achieving results in a reasonable amount of time. Such problems are associated with life sciences, biometrics, logistics, parametric database search and quantitative finance, among many other commercial and scientific areas.

As an example, consider the modeling of a nanosized structure, such as a drug molecule, using non-quantum computers. Solving the Schrodinger Equation more than doubles in difficulty for every electron in the molecule. This is called exponential scaling, and prohibits solution of the Schrodinger Equation for systems greater than about 30 electrons. A single caffeine molecule has more than 100 electrons, making it roughly 10^44 times harder to solve than a 30-electron system, which itself makes even high-end supercomputers choke.

Quantum computers are capable of solving the Schrodinger Equation with linear scaling exponentially faster and with exponentially less hardware than conventional computers. For a quantum computers, the difficulty in solving the Schrodinger Equation increases by a small, fixed amount for every electron in a system. Even very primitive quantum computers will be able to outperform supercomputers in simulating nature.

"Quantum technology delivers precise answers to problems that can only be answered today in general terms. This creates a new and much broader dimension of computer applications," Martin said.

"Digital computing delivers value in a wide range of applications to business, government and scientific users. In many cases the applications are computationally simple and in others accuracy is forfeited for getting adequate solutions in a reasonable amount of time. Both of these cases will maintain the status quo and continue their use of classical digital systems," he said.

"It's rational to assume that quantum computers will always contain a digital computing element thereby increasing the amortization of investments already made while expediting the availability of the power of quantum acceleration," he said.

For more technical information quantum computing, read D-Wave founder and CTO Geordie Rose’s blog.



Comments     Threshold


This article is over a month old, voting and posting comments is disabled

What they don't tell you
By darkfoon on 2/14/2007 7:03:17 PM , Rating: 2
Computer encryption algorithms (especially RSA) rely upon the assumption that their algorithms take polynomial time to brute-force an encryption key.

However, since this quantum computer can solve NP-complete(and IIRC, most encryption algos are NP-complete) problems in linear time, the time it takes to brute-force an encryption key becomes significantly less.
You can bet Governments will be buying these machines to break encrypted data.

Privacy is gone for good!




RE: What they don't tell you
By KristopherKubicki (blog) on 2/14/2007 7:08:07 PM , Rating: 2
I'm of the school of thought that our government has had machines like these for years.


RE: What they don't tell you
By brenatevi on 2/14/2007 7:31:24 PM , Rating: 2
There are ways of keeping your privacy. You just have to avoid computers. =P


RE: What they don't tell you
By darkfoon on 2/14/2007 7:50:54 PM , Rating: 2
I would believe that they have more supercomputing power than they lead people to believe, but I do not think they have machines like this.

I remember reading an interview with Jim Christy, where he states that he is "worried" that he and his team won't be able to crack a drive with Vista's Bit-locker Drive encryption on it. Now, that could just be reverse psychology and in reality bit-locker could be easier for his team to break than competing technologies.

So perhaps, there still is a need for faster computing that the gov't doesn't already have, and cracking encrypted volumes still takes considerable time.


RE: What they don't tell you
By KristopherKubicki (blog) on 2/14/2007 8:11:12 PM , Rating: 3
Well, let's consider two things: 1.) Christy is an exec in the Air Force investigation unit. I'm not knocking the AF, it certainly is the most *wired* department in the military, but there is a lot of evidence that suggests the capabilities of the NSA and CIA are significantly more advanced.

2.) In the late 80s and early 90s, U.S. private research groups and academia had a golden age with regards to quantum algorithms. Heck, everything Shor touched was gold research at the time. Yet from 1995 to 2005, the research goes a little dark. You still have announcements and breakthroughs, but its once a year instead of every three weeks. Since 2005, you have a new announcement every week about quantum computing again.

Not to sound like a conspiracy nut, but if a government agency had a quantum computer, it would be the 21st century equivalent of "the bomb." How many other research projects has the U.S. government easily hid for decades before the public caught up? (hint, it's really easy to do when you're fighting a war).

I think its entirely possible the U.S. has been doing research on this outside the public eye for decades, and a few prototype machines tucked away for high-priority uses seems very plausible.


RE: What they don't tell you
By msva124 on 2/14/2007 8:24:34 PM , Rating: 2
Nobody knows for sure. I'd say it's unlikely, given that back during WWII it was "in" to be a scientist for the government, cause you were saving the world. Now they're kind of at odds with the scientific community.


RE: What they don't tell you
By Ringold on 2/14/2007 8:29:59 PM , Rating: 2
Well, using the NRO example, it started in 1960 but its very existance was classified until 1992.

Like I said in another post, I think it's wrong to underestimate scientists in secret facilities with billion dollar budgets.

Likewise, I fully believe that the discussion is 99% moot because if they do have it it's likely that few people breathing today will be around to hear them admit to it. The government is becoming more opaque over time, not more transparent, so unlike some Cold War toys, things they've got today may never be disclosed.


RE: What they don't tell you
By paydirt on 2/15/2007 10:21:50 AM , Rating: 2
There's a lot of secrecy going on. For example an acquaintance of mine wrote a research paper on training sensors to behave like the human eye; the government immediately classified it and said it was theirs. This was probably more than 10 years ago.

Regardless, I prefer to remain blissfully unaware. I like my life better this way. :)


RE: What they don't tell you
By hubajube on 2/15/2007 1:32:46 PM , Rating: 2
I don't know about the CIA but the NSA has a HUGE budget and a ton of very advanced tools. Not to mention, VERY smart people.


RE: What they don't tell you
By Ringold on 2/14/2007 8:25:50 PM , Rating: 2
I'm of a similar school of thought as Kristopher there. I'm by no means a conspiracy theorist, but the fuzzy nature of the NRO budget means that only god and the individual agencies themselves have the slightest clue as to what they have and don't have. It would neither surprise me if they had a quantum computer in every spy satellite or if they didn't. And the NRO is only one of an alphabet soup of intelligence agencies that're steeped in secrecy and flush with more money than the GDP of third world countries.

As wasteful and inefficient as government is, never underestimate the power of a few PhD government wonks in secret lairs/labs and billion dollar budgets. ;)


RE: What they don't tell you
By Kougar on 2/14/2007 8:15:36 PM , Rating: 3
Intriguing thought... this did seem rather out of the blue to go from theory to finality.

I still say that one photo reminds me of the warp core of Cochran's Phoenix.


RE: What they don't tell you
By Kougar on 2/14/2007 8:17:24 PM , Rating: 2
I take that back! It's the prop from Honey I Shrunk the Kids movie!! ;)


RE: What they don't tell you
By shadowofthesun on 2/14/2007 10:07:51 PM , Rating: 3
Interesting point- however, another key part of quantum computing is quantum encription, where its theoretically impossible to break the code regardless of computing power or resources, because attempting to read the data itself fundamentally alters it. It comes from the idea that quantum-mechanics can't be observed without changing the reaction. I don't particularly understand it myself, as I don't consider myself any sort of expert in the subject. Popular Science and technology Review are two magazines I know who have published articles on the subject; I'm sure a google search would be more informative than this post. Granted, this technology is further away than the computer chip is, but it signals not the end of privacy, but the beginning of a new age of information encryption and security- makes this quite an exciting time to be alive, no?


RE: What they don't tell you
By bunnyfubbles on 2/15/2007 5:26:26 AM , Rating: 4
quote:
makes this quite an exciting time to be alive, no?


Yes, until SkyNet becomes self-aware.


RE: What they don't tell you
By mrgq912 on 2/15/2007 1:48:44 PM , Rating: 2
funny thing is just saw T3 on AMC a couple of nights ago.

sidenote: They should re-shoot T3, find a new john conner, a new super terminator, new chick, and start the movie script near the end of T3. All the BS that happens at the start of T3 is just a waste. I care about how the AI spread and how the initial resistance was started.


RE: What they don't tell you
By mezman on 2/15/2007 4:20:27 PM , Rating: 3
To continue the off topic T3 business: Did anyone else notice the fundamental flaw in Skynet's plan at the end of the movie?

Thanks to advances in computer tech, the idea for Skynet morphed from a program running on a singular server in a gov't basement somewhere in T1 and T2 to a piece of distributed software that installed itself over the internet on home PCs in T3 and talked to itself over the internet.

Then (I hope I'm not ruining it for anyone), Skynet proceeds to launch thermo-nukes to destroy it's distributed install base and destroy the network infrastructure it was using to communicate with itself.

Did anyone else have a problem with that little tidbit of writing?


RE: What they don't tell you
By angryhippy on 2/15/2007 6:00:44 PM , Rating: 2
Then (I hope I'm not ruining it for anyone), Skynet proceeds to launch thermo-nukes to destroy it's distributed install base and destroy the network infrastructure it was using to communicate with itself.

Did anyone else have a problem with that little tidbit of writing?


Yes, I thought the same thing. Not to mention how it manages to maintain itself and build construction robots, manufacturing plants, ect. Hopefully they're explain it in T4. :-)


RE: What they don't tell you
By masher2 (blog) on 2/15/2007 8:44:31 AM , Rating: 3
> "another key part of quantum computing is quantum encription, where its theoretically impossible to break the code regardless of computing power or resources, because attempting to read the data itself fundamentally alters it...."

Allow me to correct a couple things. Quantum encryption isn't "unbreakable"-- its simply impossible to eavesdrop upon without informing the receiver. But that doesn't prevent you from intercepting a message and decrypting it.

Secondly, quantum encryption has little to do with quantum computing per se, especially in the case of machines like this D-Wave prototype, which relies upon quantum tunneling, not the entanglement required for encryption.


RE: What they don't tell you
By emboss on 2/15/2007 11:17:49 AM , Rating: 2
Quantum cryptographic systems (not quantum encryption, since nothing is actually encrypted in a QM way; QM is just used for the key exchange) are unbreakable, for any reasonable definition of the word. The existance of an eavesdropper is known prior to any encrypted information being sent, so unless you're criminally negligent and transmit the message anyhow, you're fine. Eavesdropping simply becomes a DoS attack, which can be done in a much easier fashon by simply cutting the wire (err ... fiber).

Of course, it doesn't stand up to "I hide your reality and substitute my own" MITM attacks either, but since it's impossible to make a system that resists this it doesn't really matter ...

Finally, for completeness, quantum key exchange can be done without entanglement and just using polarizers. Not quite as quantum-funky as using entangled photons, but so far looks to be a lot more practical.


RE: What they don't tell you
By masher2 (blog) on 2/15/2007 11:43:12 AM , Rating: 2
> "Quantum cryptographic systems...are unbreakable"

No, that's the wrong way to view it. We already have an unbreakable cryptographic cipher...the One Time Pad (OTP). The problem is, of course, key distribution, which makes it impractical except in rare cases.

QC is all about key exchange...the only "unbreakable" part is the OTP we already know and use. QC allows keys to be exchanged without fear of 'surreptitious' eavesdropping, which makes OTP much more practical.

However, OTP + QC is only secure if sender and receiver have a means to validate each other. If not, then a MITM attack can still succeed, as you point out. Saying this "doesn't matter" because no system has (as of yet) been unable to resist such an attack just shuffles the blame up a level.

The reality is that, while QC certainly promises a huge step forward in cryptography, its far from a "magic bullet" that will solve all our cryptographic needs.


RE: What they don't tell you
By hubajube on 2/15/2007 1:36:31 PM , Rating: 2
Jesus masher2!!!! You are very well versed in many subjects. Spend your off time buried in books, huh?


RE: What they don't tell you
By emboss on 2/15/2007 1:44:28 PM , Rating: 2
"No, that's the wrong way to view it. We already have an unbreakable cryptographic cipher...the One Time Pad (OTP). The problem is, of course, key distribution, which makes it impractical except in rare cases."

Which is exactly why I used the word "system".

"Saying this "doesn't matter" because no system has (as of yet) been unable to resist such an attack just shuffles the blame up a level."

No, it's acknowledging basic information theoretic limitations. If there is a MITM intercepting all channels and there is no pre-arranged way of verifying the person at the other end, it is simply impossible to have a secure channel. Since this is a limitation of any system, it doesn't matter when talking about the security of a particular system (such as a QKD-based system).

"The reality is that, while QC certainly promises a huge step forward in cryptography, its far from a "magic bullet" that will solve all our cryptographic needs."

The most obvious problem being that QKD isn't really much use for encrypting stored data ...


RE: What they don't tell you
By masher2 (blog) on 2/15/2007 2:02:37 PM , Rating: 4
> "No, it's acknowledging basic information theoretic limitations...

There is no rigorous proof of this. Can I think of a way to prevent such MITM attacks against symmetric encryption? No...but a couple decades ago, most experts would have considered eavesdrop-proof communications to be impossible as well.

And I have to point out that asymmetric cryptography is inherently immune to such an attack. So, in the case where an attacker has the ability to intercept all channels and pose as any user, a public-key system would be more secure than a quantum-based solution.

> "Which is exactly why I used the word "system"...

The point though is that quantum entanglement doesn't make the encryption unbreakable. The OTP does that. All QC gives us is security from surreptitious eavesdropping.


RE: What they don't tell you
By Murst on 2/15/2007 10:05:10 AM , Rating: 2
Well, if you want to understand more about what you've posted about, here's a link!

http://en.wikipedia.org/wiki/Uncertainty_principle


RE: What they don't tell you
By senbassador on 2/18/2007 2:03:33 PM , Rating: 2
But wouldn't it be relatively easy to just copy the data into a regular-bit based buffer and work with that?


RE: What they don't tell you
By Kuroyama on 2/14/2007 11:52:52 PM , Rating: 2
The difficulty of breaking RSA is based on the difficulty of factoring, and it is not known if factoring is NP-hard or not. My understanding is that quantum computers have only been shown to be better for very specific types of problems, often where some Fourier approach can be applied, and in particular I do not think they have been shown to solve any NP-hard problems. But I may be mistaken.


RE: What they don't tell you
By smitty3268 on 2/15/2007 12:48:18 AM , Rating: 4
Quantum computers can only solve BQP problems, which are definitely NP-hard but not necessarily NP-complete. However, they can definitely factor numbers using Shor's algorithm in O((log N)^3) time and O(log N) space. I believe it needs just over 2N qubits to solve N-bit RSA, so this 16-qubit system could only work on 8-bit RSA. They're claiming to have 1000 qubit systems by the end of 2008, but I'm skeptical. They even admit they're not sure the same principals will work as they scale the number of qubits up.

From Wikipedia:
quote:
BQP is suspected to be disjoint from NP-complete and a strict superset of P, but that is not known. Both integer factorization and discrete log are in BQP. Both of these problems are NP problems suspected to be outside BPP, and hence outside P. Both are suspected to not be NP-complete. There is a common misconception that quantum computers can solve NP-complete problems in polynomial time. That is not known to be true, and is generally suspected to be false.


RE: What they don't tell you
By peternelson on 2/17/2007 7:44:44 AM , Rating: 2
Factoring is useful for prime number studies, decryption etc.

Quantum computing may help us in these as mentioned in "Prime Numbers: A computational perspective" text.

IBM have already built a Quantum computer based on 4 Qubits.

It works.

They gave it a problem: take the number 15 and factorise it.

Very quickly the machine gave the solution that 3 and 5 are factors of 15.

That would be great to know if I or any child could not do it in their head.

Now, 8 qubits is obviously better, but we need a lot more quantum bits to rival the size numbers already being factorised by traditional methods like brute force or optimised methods like number field sieve.

Additionally, from the other daily tech article: "an examination into the technical details of Orion reveals that it is not a true quantum computer in the traditional sense of the term. D-Wave Chief Executive Herb Martin said that the Orion is not a true quantum computer, but rather a special-purpose machine that uses quantum mechanics to solve problems."

Nonetheless if it is COMMERCIAL (unlike the IBM one yet) CAN I BUY ONE? If not it's not really a "hard launch" is it?

If so HOW MUCH DOES IT COST?
How much extra does the refridgeration equipment cost and running costs?

If they can get the number of qubits up to a useful level, I definitely have uses for this technology, but an 8 qubit machine is just a proof of concept toy.



RE: What they don't tell you
By encryptkeeper on 2/16/2007 3:25:01 PM , Rating: 2
Yes, but can it play Crysis at full resolution?


RE: What they don't tell you
By senbassador on 2/18/2007 2:07:39 PM , Rating: 2
But couldn't you just rely on "non-technological" methods of encryption. For instance, use code words instead of actual words, use a different language or even make up your own.

Just remember that during WW2, we used Native Americans to speak in their native language, a very obscure language, that couldn't get broken mathematically by the Germans.


"Nowadays you can buy a CPU cheaper than the CPU fan." -- Unnamed AMD executive











botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki