Scientists step closer to quantum computer chips

 

CUDOS builds microscale on-chip qubit emitter.

Australian researchers have developed an on-chip, microscale photon pair emitter that could yield more complex devices for future quantum applications.

The research was part of the Centre of Excellence for Ultrahigh Bandwidth Devices for Optical Systems’ (CUDOS) Quantum Integrated Photonics project, which aimed to develop a chip capable of both generating photons and performing logic operations within five years.

With collaborators from the UK and France, the CUDOS team demonstrated an 80-micron-long emitter of correlated photon pairs – 100 times smaller than devices used by other groups.

CUDOS director Ben Eggleton explained that the device could be one way of generating quantum bits (qubits) for future quantum cryptographic and computing applications.

“There’s an international race to create the building blocks of these quantum technologies,” Eggleton told iTnews.

“Sure, it’s still a long way from application ... [but] it’s an important step to quantum information systems.”

The so-called ‘photonic crystal waveguide’ was fabricated on a silicon chip and expected to be scalable and compatible with current manufacturing techniques.

According to the team’s research paper, to be presented in Baltimore, USA next week, photons generated by the device would be routed for use in quantum logic gates or communication protocols.

Eggleton explained that the device took a statistical approach to qubit generation, in contrast to smaller, single-photon-generating quantum dots.

Its size was achieved by slowing light that passed through the device by a factor of 20 to 30, and thus increasing the non-linear effects of the waveguide.

Researchers expected the size of the device to allow hundreds to be incorporated into a single chip.

The next step for enabling its use in ‘unhackable’ quantum key distribution networks was to create entangled photons, Eggleton said.

While the quantum information market was “not a huge one at the moment”, he said CUDOS researchers would collaborate with organisations like the Sydney University’s upcoming Institute of Nanoscience on fabrication and manufacturing techniques.

Eggleton also hoped to combine the emitter with research by University of Bristol professor Jeremy O’Brien, who in 2009 demonstrated a quantum logic circuit that could factorise the number 15 (pdf).

“It scales incredibly quickly,” he said of O’Brien’s work, noting also that the three-month-old CUDOS project was on track to deliver the “world’s first chip-based quantum information device” within five years.

Copyright © iTnews.com.au . All rights reserved.


Scientists step closer to quantum computer chips
CUDOS researcher Chunle Xiong holding the new (small) and old chip (big), Ben Eggleton, and Christian Grillet.
"@peterhau: "the reason that two-factor and three-factor authentication continues to be sold, bought and utilised" That's only used for authorisation not for encryption, the underlying RSA will ..."
By MerariSchroeder
 
 
 
Comments: 10
MerariSchroeder
Apr 28, 2011 4:40 PM
Are people still thinking that the RSA encrypted internet is still secure? How long until all of the data generated today is decrypted by a Quantum computer of tomorrow? 10 years? 5 years?
Ace
Apr 29, 2011 2:40 AM
For 99% of data, it doesn't matter @MS. It's not a matter of 'safe', it's a matter of 'safe enough'. When hackers start investing in quantum computers, I'm guessing we'll all have quantum computers at home, and encryption will have increased to a new 'safe enough' level.
MerariSchroeder
Apr 29, 2011 9:36 AM
I don't think you get it. It's data from today and in the past that's at risk, as well as right up until an impervious PKI system is invented and implemented, which is not secure. You are secure today, however all one needs to do is capture network traffic today, and then be patient, they will get to know a lot of information in a few short years.

It may be the government who get the first Quantum computer, however it won't take long for strong interest for commercialisation.

BTW, Quantum encryption is already available today, the U.S. government already use it and many bigger corporations - you don't need a quantum computer for that. However it is not a mechanism for PKI. Quantum encryption is symmetric OTP encryption, not asymmetric. AES encryption is not known to have a weakness, and quantum encryption cannot break it, it can only discover the key protected by asymmetric means. The use of quantum encryption therefore, is only a pursuit to have abosulute 100% certainty about the symmetric cypher exchanges, is gives no certainty about the asymmetric exchanges.
Ace
Apr 29, 2011 11:11 AM
You're letting your paranoia get the best of you @MS. Most encrypted internet traffic is useless within 12 months - I'm assuming identity theft is your main concern. For stuff that is actually secret, then the internet is a lousy transport mechanism, and I think most organisations/companies/govt realise that.
MerariSchroeder
Apr 29, 2011 6:34 PM
Ace wrote:
You're letting your paranoia get the best of you @MS. Most encrypted internet traffic is useless within 12 months - I'm assuming identity theft is your main concern. For stuff that is actually secret, then the internet is a lousy transport mechanism, and I think most organisations/companies/govt realise that.


Where did this idea of paranoia come into it. Oh, you realised I was right and are trying to smear me. I don't care about the implications. My point is that current transmissions are not secure. Whether the data is relevant or not in 5 years is a completely separate issue. For individuals in Australia it might seem benign. But considering many people don't ever change their password, identity theft would be very easy. and many people live in politically oppressed situations, who was it that slander the president? Who was it that leaked document x?

The public have a right to know that a future quantum computer, may be used to crack internet traffic today. Whatever that content is.
Ace
Apr 30, 2011 12:43 AM
You don't have to be living in a politically oppressive country to have your internet and voice traffic closely monitored. For your information @MS, no encrypted transmission is ever secure for all time. As as I keep saying, it is not necessary in any case.
MerariSchroeder
May 5, 2011 2:51 PM
Ace, you don't appear to understand cryptography in full. I have been referring to current asymmetric ciphers. Shor's algorithm is only effective against the likes of RSA and ECC. So all I'm saying is that people should be aware of this, why do you disagree?

Ace wrote: "no encrypted transmission is ever secure for all time"

Symmetric algorithms have been found with flaws, however AES-256 has never been broken. And is considered unbreakable - do yourself a favour and look into it a bit more : [http://en.wikipedia.org/wiki/Advanced_Encryption_Standard]. AES requires 2^256 brute force cycles - that's around 1 with 77 zeros. There is a vast difference between "beyond the end of the universe" for AES (symmetrical) and the very possible 5 years for RSA (asymmetrical).

And don't side track the debate by talking about some magical fast computer in the future. You would need to get the brute-force processing time to less than a year to be practical. Even if you could test a key per clock cycle, and had one trillion cores, you would need a clock rate of slightly more than 300,000 million trillion trillion trillion GHz on each core (that's 3.6717430630808027468154168254911e+47 GHz)! And that would only unlock one key, of the multiples used every day in short web sessions (distributed by the soon-to-be-broken RSA).

Edited by merarischroeder: 5/5/2011 02:52:31 PM - Added bold emphasis, and addressed Ace at the beginning.

Edited by merarischroeder: 5/5/2011 02:54:09 PM
realitybites
May 5, 2011 4:53 PM
"and had one trillion cores, you would need a clock rate of slightly more than 300,000 million trillion trillion trillion GHz on each core"

/me does best Little Britan voice -- I want that one!
peterhau
May 6, 2011 11:00 AM
@merarischroeder: the reason that two-factor and three-factor authentication continues to be sold, bought and utilised is that there is the unknown quotient in the current environment. You could crack the algorithm in a predicted timeframe of "5 years", and know what the token is generating, but then you have to identify the user ID, or, in the case of three factor, the truprint id as well. by the time you had worked all of this out, the data is already replaced.

this is not what the people who cracked the RSA security layer were after. they cannot extract data without the other components in the authorisation process.

I use RSA, will continue to use RSA and it seems that most of my clients will as well.
MerariSchroeder
May 6, 2011 3:16 PM
@peterhau: "the reason that two-factor and three-factor authentication continues to be sold, bought and utilised"
That's only used for authorisation not for encryption, the underlying RSA will still be cracked. This is not about people logging in as someone else, it's about all the data being intercepted and decrypted.

@peterhau: "this is not what the people who cracked the RSA security layer were after. they cannot extract data without the other components in the authorisation process."

You're wrong. Please research cryptography further, particularly TLS/SSL.

RSA (and other asymmetric) is used to both sign and encrypt. In encryption it is used to initiate symmetric encryption. RSA for example is used to encrypt the symmetric keys. On the web under TLS, the client would basically choose a new random symmetric key and then use the public web servers public key to encrypt that symmetric key. The webserver would use its' private key to decrypt the message and therefore have received the symmetric key securely.

If one records traffic flowing through their network (perhaps at an ISP). With a Quantum Computer, one may discover the private key of an asymmetric keyset, and therefore decrypt the symmetric key at the start of an SSL session. With the symmetric key, all of the following ciphers can be quite easily decrypted.

Even though encryption occurs with symmetric cryptography (AES), the keys are distributed with asymmetric (RSA), and therefore the whole process is very easy to break. The weakest point in the chain is asymmetric, and it breaks the whole public web security system.

Edited by merarischroeder: 1/6/2011 01:52:39 PM
Comments have been disabled for this article.
 
 
Top Stories
Photos: Highlights from SAP Sapphire Now 2013
All the keynote action from one of the world's biggest SAP events.
 
How do I: Improve my presentation skills
A repeatable process to follow.
 
Photos: NextDC builds S1 data centre
Prepares for September launch.
 
 
CUDOS researcher Chunle Xiong holding the new (small) and old chip (big), Ben Eggleton, and Christian Grillet.
Sign up to receive iTnews email bulletins
   FOLLOW US...

Latest VideosSee all videos »

Latest Comments
Polls
Do you prefer the Coalition's NBN policy?

   |   View results
Yes
  19%
 
No
  81%
TOTAL VOTES: 1613

Vote