[Beowulf] The Value Cluster

[Jim Lux]

----- Original Message -----
From: "Douglas Eadline, Cluster World Magazine" <deadline at linux-mag.com>
To: "Jim Lux" <James.P.Lux at jpl.nasa.gov>
Cc: "Beowulf Mailing List" <beowulf at beowulf.org>
Sent: Friday, December 17, 2004 5:57 PM
Subject: Re: [Beowulf] The Value Cluster


> On Fri, 17 Dec 2004, Jim Lux wrote:
>
> > A year or so back, there was a discussion on the list about building a
> > cluster entirely of stuff bought from WalMart (based on the $199
computers
> > Walmart was selling back then).
>
>
> The Walmart systems are what started this idea. I had purchased a $199 box
> with a VIA C3, then some Duron boxes. The problem with the
> "Walmart/Microtel" approach was that the consistency of the boxes was not
> very good. Indeed, several times I got something better than advertised!
> Motherboards would change, larger hard-drives, slightly faster processor,
> etc. So it could be hard building a reproducible cluster. Plus the cases
> they were using were vented on sides -- not good for close stacking. So we
> decided it would be better to use parts that were available from several
> sources so our efforts could be easily duplicated.

My Microtel is somewhat noisy, as well.

In your $2500 cluster challenge you laid out some requirements, but didn't
have something like "minimum number of processors"...  Is a cluster of two
legal?  I know you're using 8, which seems a reasonable number to
demonstrate lots of things (and you certainly run into scaling problems by
that time).

Do they have to be in cases?  Think of the several clusters made of bare
mobos stacked on a shelf, or threaded rod, or,....  Save yourself $50/case,
and on 8 cases, you might buy another processor?
Since the fundamental purpose of a $2500 cluster is pedagogy, it's
instructive to go through this tradeoff for all scales, from 4 processors to
1024.

One suggestion I have is to assign some value or limit to "fooling around
time", just to keep the concept of the dirt-cheap cluster sound.

Jim Lux