Dan, I would find it difficult to retire this evening unless I respond to you and bring
this to closure. No, no one except for Jim Kayne replied to the ascendant question. I
guess everyone is too consumed over in the Gilder thread. Some fairly technical
stuff going on over there, I noticed.
BTW, a favor. Let's not further complicate this discussion by the use of the
P- word. Thanks.
I could have phrased my words better wrt: "getting back at.." Did I actually say that?
In any event, there was no condescension or snippiness intended, rather simply
pointing out that there were some very interesting dynamics going on. .
The most recent additions to TERN's line card of products was expressly "not" my
focus in the post you originally responded to, and I stated such. Instead, the major
point behind my diatribe was to contest the labeling of s-cdma as an ascendant
technology.
The reason for this is that there is nothing unique about the class of device known as
cable modem, of which TERN is a member, and within whose Cable Labs standards
everyone gets a crew cut the first day in. (Military talk) TERN has a lot of peers who
attempt to solve the same problem, not only in the tree and branch model where it
finds itself, but also in its future voip and streaming video niches at the higher layers in
the stack. But here is where the interpretation of what "ascendant" actually means
becomes crucial. I could be wrong, depending on one's interpretation, in other words.
The mere fact that it "does" have plenty of company makes it a "me too," which in my
book disqualifies a candidate from being classified as an ascendant. Drilling down
further, the S-CDMA attribute is a lot of smoke when it's used "as a singular
justification" for such a lofty standing.
You might recall, btw, that I did allow that TERN's most recent acquisitions, which
you've alluded to, might be responsible for improving their outlook and
strategic direction, but that those issues were not on my hit list in my original message.
The single most objectionable issue to me remains the notion that the company
should be regarded as especially superior, or ascendant, on the merits of that one
protocol. Now, if I was reacting without due cause because the company itself is
being regarded as ascendant, and not the s-cdma "technology alone," then I yield and
offer no contest. I know it's a very picky semantic matter that I'm dissecting here, but
look at the implications of the moniker, and what lies in the balance because of its
application. If fewer cable plants get upgraded because of its heightened popularity
due to the label, then it effectively retards optical progress in the networks in which it
is employed.
I've read in many places that the s-cdma modem could be used in HFC too. I'd have
to ask "why"? If they use the s-cdma variant for coaxial systems in order to beat the
noise problems in the upstream, and if they also have DOCSIS compliant modems
for HFC, then why would they want to use the s-cdma variant for HFC? Here, I
admit ignorance, if there is a rational explanation for this. If so, please tell me what it
is. I submit that their primary intended use of s-cdma is for making all coaxial analog
networks data ready. And if they use these for HFC, then something is redundant,
somewhere, and some feature set is not being used optimally.
By classifying a company as ascendant, it skyrockets to the moon. IMO, the s-cdma
is retrogressive, not ascendant, and the other technologies they've acquired, while
they may be superior in their own right, are only one of many in each of their fields.
There is nothing unique about them except, perhaps, the source code, which is
written to do the same things that many other products unique software is supposed
to do. Am I slamming them? Hell no. I'm simply not going to place them on a
pedestal for doing what others in their space are doing.
In fact, s-cdma is best used when the decision has been reached by the operator, for
whatever reason (I've covered those reasons in my original post), to avoid building
out their networks with fiber. Which, IMO, is as retrogressive as the ILECs' placing
their bets on copper only flavors of dsl deployments which will lock them into very
long term cupricity, discouraging any hopes of ever seeing fiber in those serving areas
for many years out. The plain English implications of this is long term 256 k, or 640 k,
as opposed to eventual GbEthernet. Same goes for the classic cable modem folks,
btw. And that includes all of them today, although at first the ceiling might be higher,
eventually it will be as stated.
We've got things happening here in several different dimensions. On the one hand,
you are assessing the company's ability to penetrate at the MSO level, and if they do
that successfully, then they will have proved their ascendance.
On the other hand, I view "what" they are penetrating those operators "with" as being
more important, and not necessarily how many units they sell at this time. To me, the
fact that what they and their competitors are doing is not migratable to the
next logical level, i.e., a purely digital baseband modality, makes them all equally
unacceptable for meeting future full service networking needs in an optimal manner.
Yes, full service nets will evolve and include cable modems. We're seeing those now
in cable, but these will remain bifurcated architectures [analog and digital] as long as
the video component is NTSC analog, through and through, and the data component
is analog-to-digital. What will be needed, instead, is a migration path to pure digital.
And that means having a plan in hand to eventually trash (the sooner the better,
instead of adding more... ) r-f gear, and replacing the head end CMTS stuff with
Internet routers and IEEE LAN switches. I know, it's not that simple. There is the
fiber count in the field, developing a wdm strategy, a new compression framework,
etc. etc. But it has to start somewhere.
You may want to keep in mind that in my original assessment of their cornerstone
protocol [s-cdma] claims, I stated that the other vendors' models were no better from
a fundamental architectural view, because all of them are keyed to extending the life
of the analog motif.
I promised myself that I would not go line for line on this, and at this point I really
don't think that it is necessary to do so. I recognize the value and the origins of your
beliefs, based on your trust of what Gilder has written and I not only understand your
position, I respect it as well. I hope that you take the time to understand that my
perspective is one of architecture. I see a whole bunch of money being spent on what
I consdider to be the wrong things, often. Intermediate band aid fixes at a time when
the operators don't see the tsunami coming towards then because they are looking
straight up into blue skies.
------
The following is a superb account of the future of fiber in distribution networks.
written by a Corning official. Thanks to Kenneth Phillipps on the LMT.
telecoms-mag.com
FAC
--------------------------------
"Fibre Optic Cable: Unsung Hero"
When all the dust has settled from the excitement surrounding high-speed data
transmission, whether xDSL or wireless, will people once again remember that
fibre is the thankless workhorse that enabled the revolution in the first place?
Will fibre emerge as the successor to the broadband throne?
Malcolm Barnett
There is an old adage about being ?penny wise and dollar poor.?
Applying this to the fibre question is a bit more complicated, but it still
holds water. Carriers that waste too much time on interim broadband
access methods because they perceive fibre as being prohibitively
expensive may be overlooking the inevitability of the fibre future.
It is safe to say that service provision via fibre will initially be limited in
scope, and complementary to copper solutions for the short term,
simply on the basis of cost to operators, and ultimately to what
subscribers are willing to pay. But this equation must always be
reconciled with the constant that service demands are increasingly
driving innovation, not just the cost of technology. The fact remains
that demand for information, Net-based services and applications
continues to increase, so much so that the maximum available
transport rates are doubling approximately every two years.
In time, optical fibre will be the backbone of the information
superhighway, transporting voice, video and data to businesses,
schools, hospitals and homes. Time is therefore the variable in the
equation; it is universally agreed that the price of optical components
and cables is falling more rapidly than that of copper cables. As the
clock rate for microelectronics increases, the demands on
interconnection technology become increasingly difficult to meet using
conventional electrical solutions. As a result of this rapid growth,
electronic functions in communication networks eventually will be
replaced by photonic functions, which provide higher
information-carrying capacity.
------
continued in next post |
