[Beowulf] $2500 cluster. What it's good for?

[Andrew M.A. Cater]

On Mon, Dec 20, 2004 at 05:06:10PM -0500, Robert G. Brown wrote:
> On Mon, 20 Dec 2004, Andrew M.A. Cater wrote:
> 
> > If you have a long running problem - DON'T start it now. If it needs to
> > run for two years - buy next year's equipment (which is twice as fast as
> > today's) and run it for just one year. One year wait then one years intensive
> > compute - and you're still ahead. Next year's computer is
> > _automatically_ faster and potentially much better value for your $$ :)
> 
I _know_ this - didn't you catch the smiley above :)

Withal, this is, as ever, a useful project and discussion point. It's
_always_ worth getting a Doug Eadline / Jim Lux / rgb conversation
going - a whole lot of stuff comes out of the woodwork.  It may be
symptomatic of a new interest in clusters / better publicity for
Beowulfs or whatever but we're starting to see newcomers on the list
again after what seems a long while without too many.  As (still) a
relatively clueless cluster person, it seems to me that it is 
still the best policy to read the newsgroup and ponder deeply 
upon what you read.

Most of the answers are on this list somewhere : the only difficult 
thing to deal with is finding that someone else asked your question six 
years ago :)

All the best to all the list - whether you celebrate or not - can I wish
everyone happy holidays.

Andy