Brian D Williams wrote:
> From: firstname.lastname@example.org (Peter C. McCluskey)
> > email@example.com (firstname.lastname@example.org) writes:
> >>>Brian D. Williams said:
> >>> The future? if we ever get permission to do long distance, I
> >>>predict flat rate pricing.
> >>I believe that the customers will demand it. Why will the LDCs
> >>be able to provide unlimited service at a flat rate?
> > For the same reason that ISPs are currently able to provide
> >"unlimited" internet access. There doesn't appear to be any major
> >obstacle to switching most of the phone system to run over the
> >internet (although there are lots of minor obstacles and the
> >incentives to start the switch are fairly weak). Of course the
> >term "unlimited" as used by ISPs is somewhat deceptive, but
> >is close enough to the truth to satisfy more people than it
> Actually there IS a major problem, the Internet will basically have
> to be expanded by factors of several thousands to even begin
> handling the traffic.
> Even a small C.O. has 100,000 lines, times 64kbs of DEDICATED
> bandwidth (we call it QOS or quality of service) and your talking
> trunks of 64,000,000,000 bits per second, or 3-6 OC-192 circuits
> and the routers to match per C.O. Now your talking times 200 or so
> C.Os per state times 50 states.
multplied out you get 64gbit x 10000 = 640 terabits / sec
> My point? The existing Internet has only a tiny fraction of a
> percent of the bandwidth needed, and virtually none of the quality
> assurance based routers it will need.
A single optical cable can carry at least 6.4tbit/sec, so 100 optical cables
can provide all the required dedicated bandwidth to serve the same stuff that
all those copper cables do. The cables are there now, and are simply being
lit up as needed by the various companies running them. I'm sorry, but the
times when the data amounts used for voice seemed big are over. In the
near future the data passing over the Net will dwarf what the phone carriers
deal with now.
-- Brian Atkins Director, Singularity Institute for Artificial Intelligence http://www.singinst.org/
This archive was generated by hypermail 2b30 : Mon May 28 2001 - 09:56:20 MDT