Newsgroups: comp.parallel.pvm
From: john@elroy.jpl.nasa.gov (John Wright)
Subject: Re: Practical Parallel?
Organization: Image Analysis Systems Group, JPL
Date: 1 Apr 1996 17:22:30 GMT
Message-ID: <4jp3cm$6hd@elroy.jpl.nasa.gov>

In article <315DFA11.4DD59E43@inetnebr.com>,
Justin Hannah  <sparc@inetnebr.com> wrote:
>Well I am astounded about all the tech questions (which I hope to get
>involved in later) but I have looked over the readmes of PVM and I would
>like to know what others out there are using it for...
>
>The only application (ok, so i'm not that much of a heavyweight user) I
>would like to use it for is gcc (and all related utilities).  I would
>like to be able to use a network of unix PCs to compile kernels and
>software for BSD or Linux.  An hour here and there really adds up when
>you don't have an SGI like I used to have laying around...
>
>Basically, is it practical to take an existing unix app and make it run
>on multiple machines (any major modifications, etc) and how would you go
>about compiling it...

I developed a C++ class library on top of PVM and I used sleep and echo
to test it.  It works fine for running existing unix apps, including
make which leads right into a parallel make utility.  Basically, if you
can open 6 windows, rsh into 6 machines, run commands on the 6 machines,
and realize some improvement in performance over what you could have
done on a single machine, then PVM can do the same thing.  Piece of
cake.  Just my $0.02 worth.

John

-- 
john.r.wright@jpl.nasa.gov
http://www-dial.jpl.nasa.gov/~john

