Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!uwm.edu!uakari.primate.wisc.edu!aplcen!haven!umbc3!chimiak
From: chimiak@umbc3.UMBC.EDU (Mr. William J. Chimiak )
Newsgroups: comp.dcom.lans
Subject: Re: high speed networking between buildings
Message-ID: <2333@umbc3.UMBC.EDU>
Date: 27 Sep 89 12:15:25 GMT
References: <4574@ursa-major.SPDCC.COM> <337@ai.etl.army.mil> <2314@umbc3.UMBC.EDU> <5971@b11.ingr.com>
Reply-To: chimiak@umbc3.umbc.edu.UMBC.EDU (Mr. William J. Chimiak (MMA))
Organization: Univ of Maryland Baltimore County
Lines: 20


> Broadband with FDDI? I don't follow. Someone enlighten me (nicely).
>
> tony goodloe

I am sorry for perhaps being misleading, but it is my impression that the
nature of the question is the nagging problem that workstations now have
the I/O and processing speeds which blow past the capabilities of ETHERNET.
Folks with image processing applications and huge file transfers are irritated
with the LAN/CAN/WAN bottleneck.  One of the major problems is the media access
which roughly resides in the Data Link layer (layer 2) of the OSI model.  FDDI
is capable of delivering 100 Mbps at the MAC layer.  Besides being that fast,
FDDI has the potential of delivering real time performance - but this must
be said with a caveat.  Real time performance correctly implies an Application
Layer capability.  This is the final layer of the OSI model.  To deliver true
real time performance obviously implies that the layers three though six
deliver the same real time performance.  NIST is studyinging realtime 
performance of Layer 4 while folks like Gregg Chesson are developing the
eXpress Transport Protocol (XTP) to at least give workstations the
capability of operating on Gbps LANs.