Re: [Swarm-Modelling] Floating point arithmetic

[Russell Standish]

What's that got to do with no-one using parallel Java? Plenty of people
do parallel Fortran or parallel C/C++?

Amdahl's law is only a one dimension slice of the performance graph -
the real performance graph also includes an axis for problem
size. People typically increase their problem size as they use
faster/parallel computers. Consequently, Amdahl's law doesn't imply
parallel computers are useless (which was how it was first applied).

What I suspect is the biggest reason computational people have stayed
away from Java is the difficulty in linking to existing software
libraries written in C or Fortran (so I'm told). One of the _most_
important optimisation strategies is to link to a well tuned library
(eg Lapack or FFTW). I understand that Guava (the GNU Java compiler)
changes the picture somewhat, although people tend to stay away from
the GNU compiler for production code, as it tends to lag by a factor
of 2 of performance what the best commercial compilers can deliver. A
factor of 2 performance is a big difference when your codes run for 6
months or more!

Cheers

On Thu, May 05, 2005 at 11:11:31AM -0600, Marcus G. Daniels wrote:
> Russell Standish wrote:
> 
> >>I would have thought RMI would have been a 
> >>pretty handy feature, and I also believe there's a Java implementation 
> >>of MPI, although I haven't used either myself.
> >>   
> >>
> >
> >Perhaps, but we don't see anyone using it.
> > 
> >
> http://en.wikipedia.org/wiki/Amdahl's_law
> 
> _______________________________________________
> Modelling mailing list
> address@hidden
> http://www.swarm.org/mailman/listinfo/modelling

-- 
*PS: A number of people ask me about the attachment to my email, which
is of type "application/pgp-signature". Don't worry, it is not a
virus. It is an electronic signature, that may be used to verify this
email came from me if you have PGP or GPG installed. Otherwise, you
may safely ignore this attachment.

----------------------------------------------------------------------------
A/Prof Russell Standish                  Phone 8308 3119 (mobile)
Mathematics                                    0425 253119 (")
UNSW SYDNEY 2052                         address@hidden             
Australia                                http://parallel.hpc.unsw.edu.au/rks
            International prefix  +612, Interstate prefix 02
----------------------------------------------------------------------------

pgpKZ2bEsmYxG.pgp
Description: PGP signature