Re: Oracle RAC Stored Procedure scalability

On Thu, 03 Aug 2006 08:10:24 -0400, Serge Rielau wrote:

> Now if Oracle has solved the problem of transparent parallelization of
> logic in general without instrumentation of the language I'm sure the
> parallel Fortran compiler guys downstairs from me would love to talk to
> them :-)

Now this is a general computer science and logic problem for which there is no solution yet. Simply, algorithms that depend on the particular order of steps cannot be parallelized. Parallelizing things creates a whole new class of algorithms. For instance, you cannot do quicksort in parallel, but you can break the initial data set into several subsets, use quicksort on the chunks in parallel and then merge them together. The end result will be the same. Merge step is also impossible to do in parallel. The best way to parallelize everything is to eliminate causality. That way, we can turn the world we know into an "all your base are belong to us" type of world. I'm not sure whether that's desirable or not.

-- 
http://www.mgogala.com