[e2e] ATM

[Cannara]

Demir,

My comment is a relative one.  Today's designs for products, coming out
certainly within a year or so, are at least high multi-Gb/s boxes, aggregation
systems, etc.  I makes little sense to expend 10% overhead on ATM
cellularization now that Gb links, packet over Sonet, etc. are becoming so
cheap and processors for routing so powerful.  ATM's other behaviors and
commitments to expensive LAN interfacing are also driving it into the
background for new designs.

Alex


demir wrote:
> 
> Could you elaborate more on what the reasons are for ATM's being slow? And
> what do you mean by today's design (can you give examples)? Thank you very
> much.
> 
> Alper K. Demir
> USC, PhD student
> LA, CA
> 
> On Sun, 22 Apr 2001, Alex Cannara wrote:
> 
> > Yes, and for today's designs too.  {:o]
> >
> > Alex
> >
> > Robert Moskowitz wrote:
> > >
> > > Not to start an ATM bashing thread, but...
> >  Is ATM is too slow for today's backbone needs  :)
.