$10 Million will buy a Quantum Computer?
S. Dale Morrey
sdalemorrey at gmail.com
Fri Jun 21 16:52:12 MDT 2013
If I'm not mistaken they are claiming their current model is using 256
qubits. If I'm understanding you correctly, you're saying this thing could
break SHA256 in 1 operation or less. That's a bit scary. Wonder what it
could do on AES or RSA. Might also explain the NSA's $40million dollar
electric bill at their new datacenter.
On Fri, Jun 21, 2013 at 4:28 PM, Todd Millecam <tyggna at gmail.com> wrote:
> On Fri, Jun 21, 2013 at 1:32 AM, Levi Pearson <levipearson at gmail.com>
> On Fri, Jun 21, 2013 at 12:33 AM, S. Dale Morrey <sdalemorrey at gmail.com>
> > I guess the only 2 questions I have are.
> > What would one of these be good at doing in layman's terms, i.e. breaking
> > crypto, solving pathway problems etc?
> > And of course just to make this on topic for the group...
> > Does it run Linux?
> It's legit, I've contacted them. It runs their own custom-built OS, which
> is really just a very lightweight vector table and you treat it as a device
> on a regular computer of your choice. I think when I asked them they said
> it was a modified version of solaris. It works like a regular machine with
> shell access, and you access the quantum computer via C++ API calls that
> they provide.
> Basically, what it can do is take the hash space of a crypto algorithm in
> terms of 2^x, and it changes it to 2^(x-q) where q is the number of qubits
> that the quantum computer is built off of. Highest qubit count that I've
> seen to date was some isreali researchers could get 13 entangled pairs, so
> that'd make breaking sha256 a take the time of going through 2^51 hashes
> over 2^64, which means that this particular computer could crack a sha256
> password in about 600 days, over a regular cpu machine which would take 4.9
> million days (not a GPU-accelerated or supercomputing cluster, mind you)
> The protein folding that it can accomplish is helpful for scientific
> purposes, and the $10 million makes sense in that context, but it's again,
> research, so not necessarily practical application.
> The downside is that this machine is liquid-nitrogen cooled because it uses
> superconductors--so the $10 million is hardly representative of operating
> costs. A single machine like this would probably cost more than $100k/day
> to operate at max load.
> Todd Millecam
> PLUG: http://plug.org, #utah on irc.freenode.net
> Unsubscribe: http://plug.org/mailman/options/plug
> Don't fear the penguin.
More information about the PLUG