Fw: [Beowulf] Correct networking solution for 16-core nodes

[Greg Lindahl]

On Fri, Aug 04, 2006 at 04:36:08PM +0100, Vincent Diepeveen wrote:

> So for the replacement personnel of Greg,
> I understand that your cards can't interrupt at all.
> Users just have to wait until other messages have past the wire,
> before receiving a very short message (that for example aborts the entire 
> job).
> 
> In short if some other person on the cluster is streaming a bit at some 
> nodes,
> then you have a major latency problem.

Vincent,

I would not stick my unfortunate colleagues with this "conversation."

No, that is not what happens. Yes, some other interconnects use the
same interrupt strategy that we do. It's not a bad thing, it's a good
thing.

The most important thing in life is to know when you don't understand
something. I'm kind of bad at it myself, but you set a record.

-- greg