[Beowulf] Re: vectors vs. loops
joachim at ccrl-nece.de
Tue May 3 06:21:18 PDT 2005
Joe Landman wrote:
> With all due respect, I think Robert is correct. The majority of
> existing scientific code base is serial code with very limited (if any)
> vectorizable content, or parallelizable content. This is in large part
> due to the way people write them.
Those people obviously do not need parallel or vector processing. We don't need
to consider their codes here. If you do, the parallelization is probably harder
than the vectorization (provided there's potential for these techniques).
> The issue may be one of labeling. What you interpret as "the majority
> of scientific codes" may be very different than what I interpret as "the
> majority ..." and what Robert interprets as "the majority ...".
I did say "many", not "the majority", and gave a hint towards some interesting
recent SC papers of Oliker et.al. which support this statement. And I did not
claim that vector processing is the solution to all problems.
I don't think it makes sense to discuss this any further as we won't be able to
count all scientific and technical applications. I just wanted to give another
view of HPC as rgb or bio-informatic people have. Sharing experiences is not a
problem as long as no undue generalizations are derived.
Joachim Worringen - NEC C&C research lab St.Augustin
fon +49-2241-9252.20 - fax .99 - http://www.ccrl-nece.de
More information about the Beowulf