Beowulf - Single Board Computers? (long)

[Gerry Creager]

Bob Drzyzgula wrote:
> 
> On Mon, Sep 21, 1998 at 10:57:37AM -0400, Douglas Eadline wrote:
> > On Sun, 20 Sep 1998, Greg Lindahl wrote:
...
> > 2) I am sure that we (and others) will develop more dense
> >    packaging schemes, however, the volumes need to be there
> >    to provide a lower cost.  In addition, keep in mind that
> >    selling and supporting a unique packaging system also
> >    requires the vendor (if they do not want to get run out of
> >    town) to be able to support such a product - which in turn
> >    means the vendor has to design and test such a product -
> >    which takes time and money...
> 
> Well, yes... this is kind of what I was speaking of, although
> I may not have done a good job of sticking to the future tense
> in what I wrote. My point was to attempt to identify possible
> starting points for this kind of product development. Clearly
> one could start with a stock SBC, and if it sold like hotcakes
> maybe have the funds and the confidence to develop a more
> custom system using parts that are specifically designed for
> the job. Sort of what Exabyte did with 8mm Video cameras.
> Although I would hope the prices of a tiny Beowulf compute
> node would fall faster than Exabyte's prices did...

No, it's not going off-topic.

I just got back from the Institute of Navigation GPS meeting and one of
the vendors there had a Linux box of PC-104 cards.  Their business
includes making PC's in a box, and pretesting them... sounds like a
match.  This company was NOT familiar with all the PC-104 players, but
those they'd found and had decided were compatible, they seemed to know
well.  Let me dig thru the contact cards and I'll post an URL and
contacts.

Gerry Creager
Mapping Sciences Laboratory
Texas A&M University