Edward and all:
I agree with your analysis: I could not have answered Jay's
email better. It would be interesting to hear what the
end users will have to say about the logic of the IEEE
if and when the 802.3ae Task Force will approve a 10 GbE
standard with no 850 nm solutions in it.
I will only add a few comments to Jay's email at the end,
following his email.
Jaime E. Kardontchik
Edward Chang wrote:
> I do not see your point at all. May be you can elaborate a little bit more
> about your point.
> At initial selection, we will try to select seven proposals, then narrow
> down to four. As you said, there are only three; therefore, we need four
> more to go at the initial proposal.
> Besides, we are missing a very important point in the proposal selection
> process. We are so deeply involved in "a bit" and "a Pico second" issues,
> and not pay enough attention to the overall technology landscape.
> The short wave, 850 nm, has been the REASON for the explosion of the Gigabit
> rate LAN and I/O markets, and it will continue to play the LEADING role to
> keep this sector of markets growing. Because, 850 nm VCSEL has the unique
> characteristics, Gigabit rate performance at Megabit rate price, which no
> other techniques (FP or DFB) are dare to compete. Today, in the Gigabit
> rate, LAN and I/O transceiver market, probably over 85% is 850 nm VCSEL, and
> 10% or so for the longwave transceivers.
> It is surprise for me to find out that 850 nm VCSEL is not included as one
> of the major proposal at this point. LAN/WAN industry is a "Market Driven"
> industry -- we have to listen to the market need. All LAN, WAN, and
> internet service providers want "best performance at the lowest cost" to
> compete, and to rapidly expand in their market share. At this moment, 850
> nm VCSEL is one of the key technologies satisfys its requirement.
> When we move up to 10 GbE, 2.5 Gbps is definitely the area 850 nm VCSEL will
> ply the key role for users to provide performance and cost advantage.
> Further more, the 10 Gbps VCSEL will be available soon, which again will
> pose a formidable challenge to 1300 nm ,and 1550 nm at the short distance
> applications. As a user, I like to promote 850 nm VCSEL in 10 GbE market to
> keep overall transceiver cost down.
> If we do not include 850 nm in the proposal now, we will never have a chance
> to embrace the advantages which VCSEL can contribute to the 10 GbE market.
> In return, it will help overall 10 GbE market to expand.
> Many years ago, in ATM_Forum, a group of companies (vendors and users)
> proposed VCSEL to be included in the ATM specification. Many people opposed
> that time. Today, we should thanks to those companies who added the VCSEL
> technology to the ATM specification (and FC..etc). The LAN and I/O, Gigabit
> rate industry is dominated by 850 nm VCSEL right now.
> I hope we all will take care of our own needs, to assure all varieties of
> potential technologies being included in the proposal. As a result, each
> potential cost-effective technology will have opportunity to prove its
> superior marketing value for the 10 GbE market in years to come. For our
> future products, we should not exclude the technology, VCSEL 850 nm, which
> has been the major revenue generator for all of us. After all, we always
> reward the PERFORMERS.
> Therefore, I believe, 1.25 Gbps, VCSEL 850 nm, PAM5-4WDM proposal is a
> sensitive, smart option to be include among others.
> Edward S. Chang
> NetWorth Technologies, Inc.
> Tel: (610)292-2870
> Fax: (610)292-2872
> -----Original Message-----
> From: owner-stds-802-3-hssg-pam@xxxxxxxx
> [mailto:owner-stds-802-3-hssg-pam@xxxxxxxx]On Behalf Of
> Sent: Sunday, April 16, 2000 7:18 AM
> To: Edward Chang
> Cc: Jaime Kardontchik; PAM
> Subject: RE: 850nm-4WDM-1.25Gbaud
> In light of recent events in 10GEA, it is unreasonable to expect 7
> proposals to go forward. To continue to act as though nothing has changed
> is to act out a charade. As of today, there will be 3, and only 3, PHY's
> going forward: serial 1.3u dFB for smf to 10km, serial 1.55u DFB for smf to
> 40km, and 4 x 1.3u DFB WWDM for all multimode.
> To even bother to respond to the charade being played out on this reflector
> is to waste your time.
Your opinion only reinforces mine: O. Agazzi- K. Parhi - N. Yousefi's
"1300nm-5Gbaud" proposal - in addition to being technically challenged -
is being squeezed between two other 1300 nm proposals,
"1300nm - 10 Gbaud" from one side and "1300nm - 3.125 Gbaud"
from the other, that already have a strong support within
the 802.3ae and will surely be included in the final standard.
The "1300nm-5Gbaud" proposal does not really
have a clear justification and reason of existence. It does
not bring a clear differentiation to the 10 GbE standard, except
for its complexity. But complexity is not enough reason
for a proposal to be included in a standard.
Jaime E. Kardontchik