Hi Mike,
Isn't data traffic still doubling in the neighborhood of 100 to 150 days?
Here's a definitive white paper on the matter, written November, 2000:
research.att.com
Here's a compilation of Odlyzko's other works:
research.att.com
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
You really must learn to distinguish between bits and bucks. If your raw metric is the number of bits flying through the cloud, you may be close to right. But, think about it. One of the largest bit-burners the universe has ever seen is Napster. Where's the profit? None to be seen. This service is parasitic upon not only the recording industry, but on the pipeline services of the SPs, be they ISPs or PSTN carriers. There's a word of difference between a Napster bit and tolled voice minute on a 10 10 xxx service. The latter is paying for itself, the former is a net drain on SP EBITDA.
So, in brief, bits don't matter as a metric for network profitability, viability or future projections. What matters is EBITDA, and in that realm, what you are finding is a vast retrenchment. xDSL new subscription rate is declining. Cable modem new subscriber rate is declining. This is not in absolute numbers, but in the rate of change.
Furthermore, over the weekend, I read in one of my articles (sorry, lost the URL reference - maybe someone here can recall this?) that the actual overall number of nodes connected to the Internet in the USA declined in the most recent quarter for the first time ever. This is significant, if true, because it indicates that the Internet "fad" may well and truly be ending.
Best, Ray :) |
