[rescue] Trailing-edge compute farm looking for gainful employment

[Geoff Koehler]

Hi All,

Well, as it turns out, I have a bunch of radio telescope data that needs to
be processed.  The data is currently stored as 512x1024 2D floating point
arrays.  Processing consists of "stripping off" 1D column arrays and
searching for gaussian peaks, spikes, and pulsating signals.  Ive written
the serial code (its rough and unoptomized, but works), and I had planned to
"parrellelize" (I hate that word) the code to run on computer clusters using
MPI.  If anyone wishes to help out and has experience with or wants to learn
parallel programming, please contact me.  I can supply endless amounts of
data.................

Cheers, Geoff

Geoff Koehler
Saskatoon, Saskatchewan, Canada
Home of the SART Observatory
http://207.195.94.13/sart/

----- Original Message -----
From: <jwbirdsa at picarefy.com>
To: <rescue at sunhelp.org>
Sent: Sunday, December 30, 2001 5:20 PM
Subject: [rescue] Trailing-edge compute farm looking for gainful employment


>    Supposing one has a farm of older, relatively slower machines (Sun-2's,
> Sun-3's, early SPARCs, 386es, very small VAXen, 68k-based Macs, etc.)
> running various Unixes (mostly NetBSD), networked together and connected
> to the Net. What does one do with it?
>
>    I've been trying to think of some interesting, moderately useful
> distributed-computing project that they could sit and crank away at
> and haven't come up with much of anything. All the distributed projects
> that I know of are distributed because even fast machines aren't enough by
> themselves -- a trailing-edge farm can't make a useful contribution.
>
>    If network Tierra (an artificial-life research project) had ever come
to
> pass, that would have been a superb application for these beasts. But it
> didn't.
>
>    Ideas, anyone? Please?
>
>    --James B.
> _______________________________________________
> rescue maillist  -  rescue at sunhelp.org
> http://www.sunhelp.org/mailman/listinfo/rescue