Beowulf with Gigabit Ethernet

Bogdan Costescu bogdan.costescu at
Fri Jan 11 11:44:54 PST 2002

On Fri, 11 Jan 2002 alex at wrote:

> I think that in general point to point performance information has a
> limited value, whatever vendors might quote on their web-page. SCI might
> be performing pretty well if it comes down to just latency and bandwidth
> between two machines, but it is a ring topology. If you have
> more machines on a ring they will share that same bandwidth. 

As with all generalizations when it comes to parallel performance, there 
are cases when actually a ring topology (or 3D torus) does give results: 
when communication is done only between neighbours. CHARMM (as it was 
recently mentioned on this list) can benefit from such setup; an example 
can be found at:

Bogdan Costescu

IWR - Interdisziplinaeres Zentrum fuer Wissenschaftliches Rechnen
Universitaet Heidelberg, INF 368, D-69120 Heidelberg, GERMANY
Telephone: +49 6221 54 8869, Telefax: +49 6221 54 8868
E-mail: Bogdan.Costescu at IWR.Uni-Heidelberg.De

More information about the Beowulf mailing list