[RC5] Chloroform Used To Build Quantum Computer

gindrup at okway.okstate.edu gindrup at okway.okstate.edu
Thu May 7 21:07:08 EDT 1998


     You don't state your e-Mail address, so I can't respond to you 
     directly.
     
     "Quantum computer" is currently ill-defined.  The term has been used 
     to describe:
     1) "chemically combinatorial" problem solvers.  These are special 
     purpose computers, expressed as chemical processes, in which 
     parallelism is achieved through having a mole or so attempted 
     solutions generated at a time.  Compared to the problem domains in 
     which this method has been tried, a mole is a pretty big number.  
     This is a purely nondeterministic parallel computation.  The 
     success/failure and mechanism of the computation are tightly coupled 
     to the quantum mechanical properties of the molecules involved.  The 
     University of Southern California has a nontrivial collection of 
     people who do this sort of thing:
     http://www-hto.usc.edu/
     
     2) "nonlocal effect" optimization finders.  This relies on the sort 
     of global information effects that appear to affect, for instance, 
     the exact selection of condensation sites in crystals.  This 
     definition is not too widely used because it is very difficult to 
     implement the solutions in this hardware.  For example, how do 
     energy minimizing aperiodic crystals form out of a tremendous 
     sequence of condensation events?
     
     3) "quantum multi-valued" parallelization is sometimes used to 
     search multiple branches in a search tree simulatneously.  If you 
     are doing this chemically, then you are doint (1) above.  If you are 
     doing this with single atoms and binding sites, then you are doing 
     (2) above.  If you are using intermediate molecules and relying on 
     superpositions of multiple states to allow simultaneous searching 
     through problem domain hunks that are represented by the multiple 
     states, then you are doing this third kind of computing.
     
     (3) is, theoretically, the most promising.  (1) requires more and 
     more reactant and the *probability* of finding the optimal solution 
     is dependent on the number of active moleclues, the search space 
     size, and the reaction predelictions.  (2) is too hard to use to set 
     up problems.  Essentially the only problems that can be solved by a 
     type (2) computer are the problems that the computing machinery 
     itself presents.  This is not particularly helpful.
     
     Type (3) computers have the advantage that, in principle, 
     parallelization like that in type (1) can be increased a few orders 
     of magnitude by using smaller data-representation machinery.  In 
     fact, if supersposition is used, the single data items are 
     represented in fractional elementary particles.
        One English group has used these thoughts to encode about 8 bits 
     in the electron of a Hydrogen atom.  they estimate that ~100 bits 
     could be usefully stored and retrieved in an H atom, so imagine two 
     RC5-56 keys being encoded on a single atom.  Then apply the "RC5 
     decoding" transformation to the combined system of a microgram of H 
     in about a microsecond.  Well, that's one challenge problem down.
        The superposition method is even more interesting than that...  
     Encode two different trial keys on one H atom by putting the atom in 
     a superposition of the two representational states.  Apply the 
     transforming physical process (which in this case is much less 
     delicate than in the previous paragraph).  Now (I tread lightly on 
     the "hard part"...) extract the answer.
     
     In principle, type (3) quantum computers can compute about 4 orders 
     of magnitude more parallel-ly than type (1) computers.
     
     Related to this is a recent result concerning the "square-root of 
     NOT".  A quantum-mechanical process that takes atoms encoding bits 
     and puts them in a superposed state.  It takes the atoms in the 
     superposed state and finishes the migration to their negation.  So, 
     when applied twice, the operator is a NOT.  It turns out that this 
     one operator is a complete set of Boolean operators.
     
     All the elements exist but for the ones that are interrupting 
     silicon computing right now...  It's too hard to get the data into 
     and out of the computing cores.  The supercomputer that was going to 
     be used to do DES-II all at once had the same problem.  It's running 
     time tripled due to I/O by the processors...
            -- Eric Gindrup ! gindrup at okway.okstate.edu


______________________________ Reply Separator _________________________________
Subject: Re: [RC5] Chloroform Used To Build Quantum Computer 
Author:  <rc5 at llamas.net> at SMTP
Date:    5/7/98 10:42 AM


Karl G - NOC Admin wrote:
     
Hello
     
Could anybody tell me, what a quantum computer is? Or give me an URL?
     
> Chloroform Used To Build Quantum Computer 
> 
> In a traditional binary computer each bit has a definite state--either 1 
> or 0. But in the quantum world, a bit is both 1 and 0.
     
Quantum world?
Yes quantum world, is a completely other thing, than the normal world. 
But it's absolutely nonsense, to assume a thing to be in two states at 
the same time. The trick is, that even from the exact calculation we can 
only say something about the propability of the thing to be in a special 
state. What the correct state is, we can only say, after we measured the 
system. Only by opening the box, we can say, if Schr\"odingers cat is 
dead or not. The interpretaion of the whole thing is still open. A lot 
of things about the measuring process are not yet understood.
Hope I didn't tell too much nonsense. 
Hilmar
     
-- 
  ___
{'   `}         
 (   )          
(/   \)         
(__u__)         
     
koala           http://www.phy.tu-dresden.de/~preusse/ 
from behind
--
To unsubscribe, send 'unsubscribe rc5' to majordomo at lists.distributed.net 
rc5-digest subscribers replace rc5 with rc5-digest
     
     

--
To unsubscribe, send 'unsubscribe rc5' to majordomo at lists.distributed.net
rc5-digest subscribers replace rc5 with rc5-digest



More information about the rc5 mailing list