Hi Sam,
>>Sorry for ignoring your earlier post.<<
Not to worry, I've been delinquent in responding to posts
myself, as you can see.
>>Why should it be any less chaotic that the effect of
Moore's law was on the MIPS business (a.k.a. the computer
industry)?<<
Not surprisingly, the metrics used to determine mechanical
advantage in the carrier space are not as well understood as
the benchmarks used for PCs and LAN systems. And the
approach to measurement is surely different, since global
administration and interconnections of back office systems
between the carriers dwarfs those of individual enterprises.
In communications, there are other factors revolving
around human behavioral habits, access, convenience,
integration of functions, etc., that should also be factored
in, in order to compute "gain." But, of course, they are not,
at least not as far as I know of.
For each of these more subtle advantages realized, there are
either back end processes that are sped up, or eliminated,
entirely. Should these be taken into account?
Can Moore's Law's implications be applied to the
unleashing of bandwidth through the use of DWDM?
Indeed, can they be separated and measured independently,
such as two separate phenomena? Of course, they can't. But
we tend to think in those terms, nonetheless.
What of all the SONET node elements that will be
"liberated" as a result of DWDM deployments over the next
several years? Do these node savings amount to
contributants in determining gain? What about all of the
emerging protocols in the stupid side of networking which
are really very smart? Will saving numerous and
drudgerous steps to get to your destination on the Internet
through these protocols amount to factors accountable for
gain?
>>Perhaps it is no coincidence that Equant and IBM are
both choosing the same month to unload their global
networks. Are these early data networks destined to become
mainframe-like dinosaurs as their customer bases desert
them for a dizzying array of alternative carriage
possibilities?<<
My first thoughts, too, concerning these outsourcing deals
was that the main strategists and engineering folks within
these orgs needed a breather from obsolescence, and
required some time to re-focus on the future, without the
day-to-day headaches of managing and constantly mending
their ancient legacy architectures, some of which are thirty
years in the making, so I think I know where you are coming
from.
>There is an assumption that the skyrocketing demand for
bits/mile/second caused by the net is going to make the
network providers rich, as the demand for voice made
phone companies rich. This is a fallacy. Technologies like
DWDM, LEOS, and HFC are pushing bandwidth supply
through the roof.<<
Except for the relevance of LEOs, which I think will
present some pricing barriers to wide scale acceptance, I
tend to agree with your assessment here.
There is a parallel working its way through the VoIP camp,
where supporters are attempting to achieve ever greater, or
should I say, ever more diminutive, pricing levels, trying to
outdo each other in who can charge the least to attract the
most customers. I'm told that this follows the M-Law
model. However, the operators of those businesses don't
want to hear that just yet. I think that here, too, as in the
case of the differences cited above between the metrics
used in PC vs comms, there are different dynamics at play,
or at least this will be the case for a little while longer,
lagging the multiplier rules that we've been referencing
here. It'll be interesting to see just how much pain the early
adopters can withstand before they are forced to either hand
over their accounts in a takeout, or otherwise exit in some
other fashion.
Your next statements are somewhat prophetic and
reminiscent of some other notions I've both read, and been
accountable for myself, from time to time:
>>There will be no regional "natural monopolies" as
bandwidth will ignore sovereign borders over telephone
lines, coax, fiber, and microwaves.
Competition will be intense and leapfrogging opportunities
will actually favor latecomers rather than first movers.
Technology churn will be as rampant as customer churn.
Low-cost-provider status will be key, but it too will
churn.<<
I think that the IP protocol has allowed us to take a peek at
this already, if we look closely at our day to day use of the
Internet from several perspectives. I wont get into this in
depth here, but from the use of my PC, I can now hire and
fire ISPs at will, sometimes using two of them for different
reasons.
Even when we use an ISP consistently, we can then go to
one of their peers or someone up/downstream and use one
of their discrete services, all the while never
adjusting the controls on the dashboard. This is so
commonplace today, that we don't even take notice of it
anymore, but before today's browsers/ GUIs existed, such
facility was only dreamt about.
The techs in access gear are now adaptable to
accommodate everything from 300 baud data, to
multi-Mbps VDSL, to CableModem, to VoIP with SS7, to
you name it, mostly made possible through DSPs that still
do not get the public attention they deserve, IMO. We'll
soon see programmable premises gear that allows for
attachment to diverse media types, as well, thus allowing
the selection of different media providers at will, from air
interface, to CM to DSL, etc. Talk about leverage!
>>Yes, the winners will be those who can exploit
economies of scale and scope, and form creative
partnerships with content providers and data distributors. It
is way too early to predict who those winners will be since
we are still early in the network buildout phase. The
consolidation phase will be more telling.<<
Scale? This, again, flies in the face of Toffler. He seems to
think that micro markets will emerge and rule the Earth.
Isn't this where I came in? <smile>
Regards, Frank Coluccio |
