Light-based quantum circuit does basic maths

Researchers from the University of Queensland have taken a significant step in the quest to build a quantum computer, creating a light-based quantum circuit capable of basic calculations and moving quantum computing closer to a becoming a reality.

Theoretically, quantum computers leave even today's most powerful conventional supercomputers in the dust. It has also been long known that hypothetical large scale quantum computers could find the prime roots of large composite numbers, allowing them to "crack" modern data encryption.

This additional computing power is a result of the quantum bits, or "qubits", upon which quantum computing is based. Qubits are special bits that use the quantum properties of subatomic particles to make calculations. Quantum computers take advantage of a special quantum property called "superposition", allowing one quantum computer bit to act as many.

"One qubit can be in two possible states, two qubits can be in four, three qubits in eight, and so on," explained Professor Andrew White from the University of Queensland, who works on the project. Thus in a quantum computer every additional quantum bit will double computing power.

The Queensland research forms part of an Australia-wide collaboration called the Centre for Quantum Computer Technology (CQCT), whose "nodes" at several of Australia's largest universities are researching various aspects of quantum computing.

The quantum circuit pioneered by the Queensland researchers involves using a laser to send "entangled" photons through a linear optical circuit, White explained. Using this technology the group was able to create a circuit involving four qubits, which allowed them to calculate the prime roots of fifteen, three and five -- such calculations will eventually be used to crack common data encryption keys.

The Queensland research group acknowledged that the theorised code cracking ability of quantum computers may be why Australian quantum computer research is in part funded by a US government defence intelligence agency, the Defense Advanced Research Projects Agency (DARPA).

White speculated on the future implication of using quantum computers for encryption cracking: "If you're using the current technology for sending information, and you want your information to be private 30 years from now, I would be very worried by this," he said.

Quantum computers are largely an experimental technology, whose full potential may not come to fruition for 20 years or more. But White remains optimistic, "we have found no reason, in principal, why they won't work."

Advertisement

Talkback 7 comments

    qubits Anonymous -- 16/12/07

    maybe i am wrong but......
    one qubit can have 4 different states
    two qubits can have 16 different states
    and so on....
    because
    one normal bit can have two different states
    two normal bit can have four different states......
    somebody seems to know really nothing about quantuum physics here -.-'

    Qubits Anonymous -- 18/12/07 (in reply to #320091752)

    Qubits have 3 possible states. 1, 0 , or both.
    As the post 2 down from here says:
    One qubit can have two different states *simultaneously*, two qubits can be used to represent four distinct numbers *simultaneously* and so on.

    15 = 3*5 Anonymous -- 16/12/07

    You mean it calculated primes of 15, *which are* 3 and 5 right?

    You make it sound as if it factorized 3 and 5 - not even a quantum computer can do that :)

    qubits Anonymous -- 16/12/07

    One qubit can have two different states *simultaneously*, two qubits can be used to represent four distinct numbers *simultaneously* and so on (e.g. 00, 01, 10, 11).

    qubits Anonymous -- 18/12/07 (in reply to #320091757)

    00, 01, 10, 11 would be correct for binary code. Would it not be better stated as: 1, 2, 3, 10. Note ‘zero’ (0) is the null, placeholder, the additive identity and it is what makes modern math.

    qubits Anonymous -- 18/12/07 (in reply to #320091880)

    One more correction, it would be 0, 1, 2, 3. because 0 is the first place holder.

    Qubits Jeremy M. Rumble -- 20/02/09

    Professor Andrew White should know that he is describing regular bits not qubits. With regular bits you double the possible combinations every time you add a bit. If I am not mistaken, quantum bits can be in four positions - 1, 0, both, and neither, therefore it would multiply the possible combinations by 4 for each qubit added.

    What would be fast is a computer that used hex instead of bin. Such a system could possibly be achieved by using chips that used varying amounts of voltage instead of on and off. I don't know if that can be done electrically or not.

    If encryption was broken by light computers, they would simply use bigger keys, and write new, better encryption algorithms.

    I know of a non feasible, but nevertheless theoretically unbreakable encryption scheme - one time pads. They aren't calculated with extremely large numbers, therefore can't be factored. They are based entirely on random numbers XOR'ed with text. Like I said though, not feasible for internet communication encryption.

Add your opinion

Latest Videos

Blogs

  • Darren Greenwood Telecom NZ savings damage prospects
    If Telecom NZ wants to have any of the NZ$1.5 billion the government intends to spend on its new broadband network, it had better think long and hard before offshoring 1500 jobs.
  • Array iiNet: The whys and what nows
    Last week the Federal Court ruled that internet service providers are not responsible for copyright violation by their customers. This is an important decision not just for iiNet, which spent around $4 million defending the case, but for all ISPs in Australia and, indeed, globally.
  • Array Govt, hurry up with releasing data
    A programmer scraped data from the My School website to make some really cool heat maps showing regions of smart schools — no thanks to the government, which didn't supply the data in any useful kind of format.
  • More blogs »

Tags

Back to top

Featured