[NTLUG:Discuss] [WTLUG:discuss] Quantum Computer and linux (fwd)

[Steve Baker]

S Jackson wrote:
> 
> The theory of quantim computing is very fascinating, but Steve is
> right.  It is way off.  Currently several simulators are being
> developed, but no quantum computer (with more than 2 qubits) has ever
> been produced.  Once you read the superposition of the particles, you
> destroy their state (this makes it difficult to get/check results).  I
> believe the reference in the original post is referring to a quantum
> simulator based on linux.

Yep.  Although I don't think the fact that it's running under Linux
is particularly significant.  This is probably some huge C program
that would run on anything.  The Quantum computer that's being
simulated won't be running any kind of OS at all.

>  If anyone is interested in this topic (it is
> actually very fascinating from a mathematical/physical standpoint) check
> out www.openqubit.org.  It's a group of coders, physicists, chemists,
> and interested scientists who are perfecting a quantum computer
> simulator, based on linux.

The radio show about this was fascinating - things like: It'll be
fundamentally impossible to use things like debuggers on Quantum
software.  This is because *ANY* attempt to observe the state
of the computer before the program finishes running will cause
the state of the machine to 'collapse' - completely screwing
up the calculations in progress.  This makes debugging not
just 'challanging' - but utterly, fundamentally, guaranteed
to be impossible by the laws of physics!

One of the people on the show remarked that one of the big
problems with Quantum computing is knowing when your program
has finished running!  Current algorithms for things like
factoring big numbers, solving the travelling salesman
problem, that kind of thing - have been crafted to run
in an exactly knowable amount of time.  However, there are
lots of things we use computers for where you can't tell
in advance how much time the program will take.  Fundamental
computational theorums state that you cannot tell in advance
(for an arbitary program) how long a program will take to
execute without actually executing it...that's a big problem
for Quantum computers.

What's really freaky is that if you made your quantum computer
light a lamp when it finished, the computer would work just
fine as long as you have no way to observe the state of the
light...but if you can possibly know the state of the light
before the computation is complete - you'll screw up the
calculation.

This is just like the Schrodingers' Cat thought experiment
done for real. For those who don't understand this stuff,
it's well worth reading up on the subject...it really
bends your brain!

I'd see these contraptions more as computational peripherals
of conventional computers rather than computers in their own
right.

-- 
Steve Baker                  http://web2.airmail.net/sjbaker1
sjbaker1 at airmail.net (home)  http://www.woodsoup.org/~sbaker
sjbaker at hti.com      (work)