On Wed, Mar 20, 2013 at 3:52 PM, Mike Archbold <[email protected]> wrote:
> Thanks.  In a kind of pure form quantum computation seems kind of
> problematic now, but it seems like it could be hybridized in the not
> to distant future.  Example:  suppose you wanted to capture all the
> properties for some object in 8 bit registers, one register to
> describe each property of the object.  Instead of being limited to a
> single property per register, now you can cram 2^8 -- 256 --properties
> in a single register.  A simply program could be crammed in another
> register, so you could run the whole shebang out of just two
> registers.  The only time I tried parallel programming was on a Tandem
> computer, and I never developed the hang of it really.  Simpler just
> to think serially.  So some means of converting a program to parallel
> from serial would be nice.  doubtless people have tried that.  Mike A

I suspect that you are thinking in terms of conventional computing.
Are all of your operations time reversible? Most familiar operations
like assignment, arithmetic, and array indexing, are not. You have to
think in terms of rotations in complex vector space. Also, you can
store a superposition of values in a qubit register, but then you can
only read one of them out, and you won't know which one.

Besides Shor's algorithm and Grover's algorithm, one application could
be computational chemistry. Currently, there is no program that can
input a chemical equation like CH4 + O2 and tell you what the reaction
products will be and how much energy will be released. In theory, you
could calculate it by modeling the motions of the atomic nuclei and
electrons. But that requires solving the Schrodinger equation, which
for n particles requires O(2^n) operations on a conventional computer.
However, nature "computes" chemical reactions with O(n) operations in
parallel in O(1) time. In theory, so could a quantum computer, if we
knew how to build one.


-- Matt Mahoney, [email protected]


-------------------------------------------
AGI
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-f452e424
Modify Your Subscription: 
https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-58d57657
Powered by Listbox: http://www.listbox.com

Reply via email to