This story appeared on Network World at

http://www.networkworld.com/columnists/2007/113007bradner.html

Internet overload: painting tomorrow something like today

By Scott Bradner

Network World , 11/30/2007

It seems like only yesterday that the press was talking about an Internet collapse, but it actually was more than a decade ago and happened because the press thought Bob Metcalfe was predicting the Internet was going to overload and collapse.

This time it happened because Nemertes Research, my fellow columnist Johna Till Johnson's company, put out a report that generated a lot of misleading headlines, including the one in Network World . Other headlines were the China Daily's “Internet outages may occur by 2010 as capacity stalls”; iTwire's “Internet to go down in 2010?"; the Times of India's “Superhighway traffic jam could clog Internet”; and InternetRetailing.net's “E-commerce could slow to a halt by 2010." That would be scary stuff if the report actually said anything like that.

The Nemertes report itself is titled “The Internet singularity, delayed: Why limits in Internet capacity will stifle innovation on the Web” and can be found on the Nemertes site. (Annoyingly, they want you to create an account to download the report; I'm going to be real pissed if I start getting spam from them now.) The report is quite well done, but still I do have some real problems with it, even though Nemertes says some of the best people I know who are thinking about these issues gave them advice, including Noel Chiappa, kc claffy and Andrew Odlyzko.

The report does not pull a Metcalfe and predict an Internet collapse. It does, however, say that the Internet broadband-access networks will not keep up with future demand and, thus, users will be slowed. It does not mention that many broadband Internet subscribers are seeing slowdowns today because of the low speed and oversubscribing of current access networks.

My second-biggest problem with the report is that it fails to take into account the wide differences in Internet-access speed and cost across the world. The average download speed in the United States is less than 2Mpbs compared with more than 60Mbps in Japan. The report fails to point out that the United States’ definition of broadband is one of the slowest and most expensive to be found in the major industrialized countries in terms of dollars per Mbps -- more than ten times as expensive as in Japan and even more expensive than in Portugal. It calls for spending a lot more money on access infrastructure over the next few years, but does not hint at how it might be paid for. Already U.S. broadband Internet service is too expensive for a lot of people, and the Nemertes report does not factor that into its projected growth in users and demand.

My biggest problem with the report, however, is that its authors seem to think that the only possible Internet-access future comes from the traditional telecommunications carriers. It ignores (or at least I could not find any mention of these) non-carrier solutions, such as muni or neighborhood Wi-Fi, and mentions Google’s potential entry into the wireless-access business only in passing.

Another report, the "Broadband Reality Check II," notes that “the U.S. broadband market is dominated by regional duopolies and little competition” with one cable and one telephone provider in each region. As long as that remains the case -- and the U.S. Federal Communications Commission seems to want to be sure that it does -- we will continue to get slow and expensive Internet service, and the more likely it will be that the Nemertes report's predictions of continued clogged access networks will be true.

Disclaimer: Harvard’s shield is all about truth, but the university has not expressed an opinion on the level of truth in the Nemertes Research report, so I’m on my own above.