RE: WWDM vs. 10Gb/s serial

[Curt Berg <cberg@xxxxxxxxxxxxxxxxxxx>]

Looks like the WAN and LAN markets definitely need 
different physical solutions. However to penetrate the
market faster, and to get ecomony of scale, it is vital that
we have one standard interface (like GMII + management),
that is independent of the physical interface. 

-Curt Berg- 
ASIC Manager, Extreme Networks

-----Original Message-----
From: BRIAN_LEMOFF@xxxxxxxxxxxxxxxxxxxxxxxxxx
[mailto:BRIAN_LEMOFF@xxxxxxxxxxxxxxxxxxxxxxxxxx]
Sent: Friday, May 07, 1999 9:11 AM
To: bill.st.arnaud@xxxxxxxxxx; stds-802-3-hssg@xxxxxxxx
Subject: RE: WWDM vs. 10Gb/s serial


     Bill,
     
     Again, you are confusing a LAN discussion with an independent, and 
     very different long-haul discussion.  I agree wholeheartedly that
10-Gb/s 
     serial, i.e. externally modulated lasers similar to OC-192 products are
the 
     optimum solution for long-haul, and they are already widely available!
     
     The SpectraLAN WWDM proposal is intended to be a low-cost solution for 
     premises and campus LANs (up to 300m on 62.5 micron fiber, up to 10km
on 
     single mode fiber).  The economics of this market, and the decision
between 
     WWDM or serial are completely separate and very different from the 
     long-haul market that you are interested in.  
     
     Regardless of which market "drives" 10-GbE, different physical
solutions 
     will be required for the long-haul and the LAN.  Nobody is going to buy
a 
     cooled, isolated, externally modulated transmitter, meeting all of the 
     rigorous specs for long-haul transmission, to go 200 meters between
wiring 
     closets in an office, or even to go 3km between buildings in a campus.
     
     Both LAN and long-haul applications are open for discussion on this 
     reflector.  It is important not to confuse the two when making
arguments 
     relating to physical layer solutions.
     
     -Brian Lemoff
     
_