Linux Adoption Barriers 3260

On Tue, 28 Feb 2006 15:03:56 -0500, Ray Ingles

From what I understand of quantum computing, the overall conceptualization of the operation is not really significantly changed; what is changed is how certain clbuttes of problems are solved.

That is, given a complex task such as, oh, prime factorization, conventional processing proceeds in a more or less linear fashion, with parallel processing simply splitting up the task into several simultaneous chunks which are handled, again, in a basically linear fashion. Contrast that to quantum computing where, to all intents and purposes, every possible solution can be tried simultaneously, such that even if the evaluation of a single "path" takes a relatively long time, the entire process takes no longer than this:

EG if it takes a quantum computer an hour to do a factorization attempt that would take a linear computer ten seconds to do, it's a net gain, as the QC manages to do *every* possible solution in that hour, whereas the linear computer can only manage one every 10 seconds - requiring, say, a billion years to complete the task.

What this suggests to me is that a QC would behave much like a conventional computer, but with something equivalent to tuples usage on parallel systems - if your task is suited for mbuttively parallel (QC-style) computation, the OS and-or libraries provide the tools by which to set up the problem a a QC solver task, launch the solver, then sleep until there's a result. Most other things - email, IM, word processing, etc, etc, etc, will still need to be done, but really don't benefit from the addition of the QC's functionality as (in essence) a mbuttively parallel system.

Given that, it would seem that the only real changes Linux would have to undergo, apart from scheduling and the like, are the addition of the QC process creation and reporting facilities.

-- MS, because work should be measured by effort, rather than result.