[Fwd: Re: 32-port gigabit switch]

Dave Vehrs davidv at aspsys.com
Fri Mar 7 07:15:16 PST 2003

On Fri, 2003-03-07 at 05:55, Jeffrey B. Layton wrote:
> Dave Vehrs wrote:
> >  
> >
> >>>Additionally the latency on Gigabit is simply to high for large
> >>>clusters.  Remember on clusters latency is more important than
> >>>throughput. 
> >>>
> >>>      
> >>>
> >>Bzzt. Wrong. Dead Wrong. "It depends" is the correct answer. (Don't
> >>pass GO, don't collect $200).
> >>    
> >>
> >
> >OK show me how to do 512 nodes or higher with Gigabit and reasonable
> >latencies (i.e "for large clusters").
> >
> It depends upon your APPLICATION. Some need bandwidth, some need
> latency (some don't really need either :) Stating that latency is more
> important than bandwidth is really just trying to start a flame war. I'm
> suprised you're trying to start a flame war =:)
> Jeff


I'm not trying to start a flame war, and I'm really curious.  I suggest
that you're starting the flame war with your attacking tone and lack of
any facts (or even one example) backing up your statements.  Just saying
"it depends" doesn't help the rest of us learn.  When is Gigabit better?

In my experience the computation portion of a Beowulf will always
require low latencies for optimal performance.  

On the other hand, when I have applications that need to transfer a lot
of data as well, I find that having two networks is the way to go.  One
for control and messaging traffic (low latency - Myrinet) and one for
data traffic (high throughput - Gigabit).  

If you would rather take it off list, then feel free to email me
directly, but I would really like to know because I can't think of one
example that works.

David E Vehrs, System Engineer		Aspen Systems
davidv at aspsys.com			3900 Youngfield Street
Tel: +01 303 431 4606			Wheat Ridge CO 80033, USA
Fax: +01 303 431 7196			http://www.aspsys.com

More information about the Beowulf mailing list