Re: tlug: Karl-Max has cool dreams [was: dual-pentium processors]

[Cliff Miller <cliff@example.com>]

> There are not many people who can write a program that
> scales to 64 processors, and if you can, it is an aweful lot
> of work.  And then there are more problems where we don't
> know any useful parallel algorithm than problems where such
> algorithms are know.  Furthermore, there is no compiler

Yes, most problems are hard to solve with parallel algorithms.

> [1] Yes, there are parallelising Fortran compilers, but try
>     to use one and you'll see that it is not much easier
>     than writing a parallel program in the first place (and
>     they only work for a small class of numercial algorithms 
>     anyway).

But, there is an important class of problems that lend themselves
very well to parallel solutions. Basically, problems where you have
to calculate the same thing over and over again on varying
inputs and then gather up the results.  Although these have typically
been seen as applying to "scientific computing", there are some
important business applications that are of this class too.

(The original point that you make is still valid -- parallelizing
sequential programs is tough, really tough, and it doesn't often buy
you a lot.)

Cliff

=============
$B%/%j%U!&%_%i!<(B
$BBeI=<hDyLr<RD9(B
$B%Q%7%U%#%C%/%O%$%F%C%/3t<02q<R(B
Cliff Miller
Pacific HiTech, Home of TurboLinux
Tel +81-3-5477-9675
Fax +81-3-5477-9919
cliff@example.com


--------------------------------------------------------------
Next Nomikai: 18 September, 19:30 Tengu TokyoEkiMae 03-3275-3691
Next Meeting: 10 October, Tokyo Station Yaesu central gate 12:30
--------------------------------------------------------------
Sponsor: PHT, makers of TurboLinux http://www.pht.co.jp