Forgot your password?
typodupeerror
Cloud Supercomputing Science Technology

Qcloud Puts Quantum Chip In the Cloud For Coders To Experiment 73

Posted by Soulskill
from the spooky-action-at-a-distance dept.
hypnosec writes "Quantum computers are currently available in very few labs, usually bankrolled by major organizations like Google and NASA. However, a new project called 'Qcloud' aims to break those barriers by making quantum computing available to everyone. The University of Bristol announced the launch of Qcloud today at the British Science Festival 2013, with the goal of making quantum computing resources available to researchers across the globe. Claimed to be the first open-access system of its kind, the quantum chip is located at the Center for Quantum Photonics at the University of Bristol. Researchers can remotely access the processor over the internet for their computational needs. Those looking to test their ideas on the processor would be required to first practice and hone their skills using an online simulator. The university has made tutorials available to researchers so they can learn how to tune the processor and change its output as required. Once they are confident in their skills, researchers can ask for permission to access the real quantum photonic chip."
This discussion has been archived. No new comments can be posted.

Qcloud Puts Quantum Chip In the Cloud For Coders To Experiment

Comments Filter:
  • by wbr1 (2538558) on Friday September 06, 2013 @09:00AM (#44774247)
    ...is that with quantum processors, you either know how to code for one, or you have access to one, but never both. :D

    That and you never know for sure if your post was first!

  • This new invention has already been pre-exploited under contract for the NSA, so any encryption derived from these techniques will still be breakable by them.

    Better luck next millennium!

    • Quantum encryption is different than quantum computing. In quantum encryption, it is impossible to 'listen' into a message without being detected.
      • Ahem... Only in theory [technologyreview.com] for now.
      • In quantum encryption, it is impossible to 'listen' into a message without being detected.

        The unintentional implication of that phrasing is that it is still possible to listen (albeit detected). What you do of course is transmit the key on the quantum channel, and if you detect anyone listening in at that point, you discard the key (actually, if I remember correctly, you just discard any intercepted bits and use the rest for the key). Once the key's got across can happily shout your encoded message from the rooftops (or other more practical classical channel).

      • which is completely useless if the NSA has a backdoor into your OS and is key logging everything you do.

        • If we assume the NSA has a backdoor to any computer system it wants, then there is no chance for any kind of privacy no matter what we do.

          I suspect this is not the case.

  • The simulator (Score:5, Informative)

    by earlzdotnet (2788729) on Friday September 06, 2013 @09:10AM (#44774341)
    I spent 5 minutes wading through news articles to finally find the free to access simulator: http://cnotmz.appspot.com/ [appspot.com]
  • I'm a doctor, not a doorstop. - EMH Mark I

  • I'm neither a physicist nor a computer scientist, but if you can "hone [your] skills" using the simulator, why isn't it sufficient to have a fast enough simulation of a quantum system using a classical computer, and solve your problems on the simulator?
    • Re:simulator (Score:4, Informative)

      by GameboyRMH (1153867) <gameboyrmh@gmaiWELTYl.com minus author> on Friday September 06, 2013 @09:49AM (#44774705) Journal

      For the same reason it's wasn't sufficient to simulate your computer programs on paper back in the early days of computing...the simulation is FAR from fast enough.

      • Only, well, they're far faster. Quantum computing is a sham and anything you can do on a "quantum computer" you can do faster on a normal computer.

        What now?

    • Scalability (Score:4, Informative)

      by hweimer (709734) on Friday September 06, 2013 @11:00AM (#44775553) Homepage

      I'm neither a physicist nor a computer scientist, but if you can "hone [your] skills" using the simulator, why isn't it sufficient to have a fast enough simulation of a quantum system using a classical computer, and solve your problems on the simulator?

      The reason is scalability. Even with the best currently existing methods, the computational complexity of simulating quantum systems on classical computers grows exponentially with the number of qubits. Quantum computers, being quantum systems themselves, do not have this exponential scaling. With just two qubits, the exponential penalty for classical simulation is rather small (and the classical simulation will be much faster), so the only reason why you would want to build an actual experiment is to make proof-of-principle tests of the technology. With a few tens of qubits, the exponential growth becomes relevant, and classical simulation becomes impractical. Right now, the world record for the full simulation of quantum systems on classical computers is 42 qubits, and the world record for quantum computing stands at 14 qubits. So, while the real experiments still have some way to go before they catch up with what we can do with classical computers, it's not crazy to think that this will happen within the next decade.

      • by Phiu-x (513322)
        This. And the holy grail is >300 qbit. I want to see this in my lifetime.
  • I don't have to test it, it already succeeded and failed at the same time.

  • by nimbius (983462) on Friday September 06, 2013 @09:54AM (#44774805) Homepage
    from the bristol site

    From the 20th of September 2013 you will also be able to access to the worldâ(TM)s first open-access quantum processor

    so its not a cloud, or even a cluster. you'll need to register for an account to use the processor and as such id expect the service is going to look more like the superdome 9000 access i had in college than anything close to cloud. FIFO or RR scheduling will be used to execute jobs simulated for time as a component of their priority level. This is actually the way every supercomputer works, we're just extending the academic luxury of such a machine to the general public.

    TL;DR: fuck your cloud, get off my lawn.

  • Awesome! Time to write a quantum Bitcoin miner - Let's see if we can waste CPU cycles by the universe itself! ;)
  • *Insert obligatory "Quantum Leap/Bakula" joke here*
  • well here's the thing...if we split a qbit and i set it to a state, before the signal from my computer tells it to set it will already be set at the lab.

    yikes talk about debugging issues.

    • Eh, people who code multithreaded programs already feel like this is happening anyway, so this is just a special instance of a race condition for those guys =P
  • Qcloud puts a chip in someone's butt, for coders to... "experiment". Fun!

    https://github.com/DaveRandom/cloud-to-butt-mozilla [github.com]

  • Just to clarify: this is a quantum processor, not a quantum chip. It is probably more room-sized than chip-sized.

    • by iris-n (1276146)

      As far as I know the Bristol folks do integrated optics, so it is in fact a chip; not as small as a classical chip, but about the size of a fingernail.

      As a side note, it makes me sad that such an awesome project had to taint itself with the mention of Google and NASA waste of millions of dollars to D-Wave. What the Bristol people are doing is quantum computing. What D-Wave is doing isn't.

  • If you want to try out quantum computing [blogspot.com] simulation, consider checking out QCF [sourceforge.net] in the Matlab-like Octave [gnu.org].

  • I swear i could see a guy in a suit.
    Pretty sure he was selling something from the the 1990's called "servers that talk to each other".

  • More like the way supercomputers are run. Youve got to submit a job to be done, and then the jobs are arranged by priority and it may take a while before your code runs on it.
    • This should still faster than running your program on a classical computer for the whole time you were in line, provided your particular problem was suitable for solving on a quantum computer in the first place.
      • No it won't. Classic computers are still faster.
        • Neither is faster or slower. Quantum computers are faster at solving certain types of problems (BQP problems), and will become many orders of magnitude faster as they are able to control entanglement between more qubits and solve larger scale version of the same sorts of problems.

Work without a vision is slavery, Vision without work is a pipe dream, But vision with work is the hope of the world.

Working...