RE: Why I think a variable date rate PHY is important for LAN or WAN
Comments:
I agree that these two issues (2-PHYs and distance) are separate issues
which we should pursue separately.
After two months of hard working by all of us in arriving two PHYs
consensus, we should nail down the PAR for two 2 PHYs in York without any
other interference. After all this is the top priority.
The installed base MM fiber is another important issue that can be optimized
without tying to the 2-PHY issue at all. They should be dealt
independently.
The utilization of the installed base is a requirement for successful 10xGbE
products. Without installed base to provide low-cost implementation, and to
support the smooth migration from 1xGbE to 10xGbE we will encounter
tremendous resistance in the marketplace to advance 10xGbE products. I
would believe for those to ignore the significance and impact of the
installed base is not realistic.
Nevertheless, the utilization of the installed MM fibers for 10xGbE were
discussed in the Montreal meeting. I presented the installed base MM fiber
proposal. For long wave, there is no doubt we can achieve 100 meter at 10
Gbps, or with the mode conditioner we can achieve 200 meters. Although some
of the bad fibers (DMD) will be discarded with TCP/IP and BER test.
If we adopt 4x2.5 Gbps approach, the distance will go to 300 meter without
any big deal.
At this level of performance, 50 um and 62.5 um fibers do not make much
difference.
If it is necessary, I will continue to provide the data and implementation
procedure to support the installed base utilization.
The new super MM fibers are good, but they will require a new installation
(cost and time), and the performance is not that much different from those
installed fibers which do not have any DMD traces. I do not see any thing
wrong to keep 100 meter, 200 meter for the time being till we know better.
As for, the shortwave (850 nm) the distance will be different, however at
4x2.5 Gbps approach, we still can reach 300 meter MM fiber with the
restricted launch transceivers. We will keep the installed base.
DO NOT GIVE UP
Edward Chang
NetWorth Technologies, Inc.
EChang@xxxxxxxxxxxxxxxx
As I see it today, we currently have 5 distance objectives, and we have a
proposal for an objective of 2 PHYs (WAN and LAN). I don't believe that the
distances have to be related to the type of PHY, and I'd prefer that we
don't do that. We are still in the phase of realizing all the possibilities
that can be incorporated into this standards effort. If someone was to
present a LAN PHY/PMD combination that could meet the 40 km distance
objective, then I would like for that to remain in the scope of our work.
It may also turn out that the PHY/PMD that meets the 100m over installed MMF
objective is the same as the PHY/PMD that meets the 300m over MMF objective.
I believe that our focus going into the York meeting has to be on creating a
list of objectives, not a list of implementations, that are realistic and
focus our scope of work, and to draft a PAR and 5 criteria. I think
Howard's proposal has created a giant leap forward for this group to be able
to close on our list of objectives and get down to the intense work of
drafting the PAR and 5 criteria.
Thanks,
Brad
-----Original Message-----
From: owner-stds-802-3-hssg@xxxxxxxxxxxxxxxxxx
[mailto:owner-stds-802-3-hssg@xxxxxxxxxxxxxxxxxx]On Behalf Of Walter
Thirion
Sent: Thursday, September 16, 1999 11:06 AM
To: 'HSSG'
Subject: RE: Why I think a variable date rate PHY is important for LAN
or WAN
Bill,
I absolutely agree that we have to support the installed base. However, with
trunking/link aggregation being standardized, you can easily trunk 2 to 4
1Gig segments to achieve the performance between 1 Gig and 10 Gig.
Considering the develoment cost, equipment deployment costs, etc., it just
doesn't make sense to me to target less than 10 Gb/s.
Walt
> -----Original Message-----
> From: Bill St. Arnaud [mailto:Bill.St.Arnaud@xxxxxxxxxx]
> Sent: Thursday, September 16, 1999 9:54 AM
> To: Bruce_Tolley@xxxxxxxx
> Cc: 'HSSG'
> Subject: Why I think a variable date rate PHY is important for LAN or
> WAN
>
>
>
> I would like to echo Bruce Tolley's comment that the
> traditional LAN market
> is important as the WAN market.
>
> This is why I think we need a variable data rate PHY ( yes I
> know this may
> be real hard with PLL). But maybe we can user
> interchangable PHY devices
> like GBIC connectors.
>
> The problem is that just like in the WAN, there will be a
> huge installed
> base of existing fiber. A lot of this fiber may not support
> 10GbE at the
> targeted distances, but it may work very well at speeds of
> 2GbE, 4GbE,
> 8GbE, etc
>
> As lasers are the single biggest cost, partularly in the WAN,
> lower data
> rates will allow us to operate our networks at greater distances. As
> bandwidth demand grows users can have a smooth upgrade path by either
> increasing laser power, or installing boosters (1R) and or
> repeaters (3R) or
> ultimately installing new fiber
>
> Bill
>
> Bill St. Arnaud
> Senior Director Network Projects
> CANARIE
> bill.st.arnaud@xxxxxxxxxx
> +1 613 785-0426
>
> Bill St. Arnaud
> Senior Director Network Projects
> CANARIE
> bill.st.arnaud@xxxxxxxxxx
> +1 613 785-0426
>
> > -----Original Message-----
> > From: owner-stds-802-3-hssg@xxxxxxxxxxxxxxxxxx
> > [mailto:owner-stds-802-3-hssg@xxxxxxxxxxxxxxxxxx]On Behalf Of
> > Bruce_Tolley@xxxxxxxx
> > Sent: September 16, 1999 8:18 AM
> > To: Young, Leonard G
> > Cc: 'HSSG'
> > Subject: RE: York plans
> >
> >
> >
> >
> >
> > Leonard:
> >
> > We cannot roll out 10 GbE in 2 or 3 years to our customers and
> > tell them SORRY
> > there is no support for your installed fiber. We CANNOT tell
> > them just pull
> > enhanced multimode fiber.
> >
> > As it stands 100 meters is not sufficient as a goal for the
> > installed base. 10
> > GbE will be a backbone technology for campus LANs and must
> > support the installed
> > cabling. If we cannot run 10 GbE on installed MM fiber at
> > backbone distances,
> > we should start telling customers now to pull SM fiber.
> >
> > What we really need is a goal that gives us a real backbone
> > distance of 200
> > meters on the installed MM fiber.
> >
> > There has been a plethora of email about support for the WAN on
> > this reflector.
> > How about some email that communicates support for our
> > traditional customer
> > base??
> >
> > Bruce Tolley
> > Manager, Business Development
> > 3Com Corporation
> >
> >
>