Broadband industry lobbyists and PR departments have ceaselessly predicted a bandwidth crunch that never seems to arrive, usually because they're trying to justify a new policy or rate hike, stave off regulation -- or in the case of hardware vendors, sell network management hardware. But buried beneath the roar of these "exaflood
" chicken littles have been more reasoned voices who claim bandwidth growth, while strong, is both reasonable and manageable. Unlike some
, these more reasoned voices usually back up their claims with real data.
Andrew Odlyzko is one of the nation's top experts on global Internet traffic. Stationed at the University of Minnesota, Odlyzko posts all of his data to his website
, and notes that while growth is strong, it doesn't necessitate drastic new pricing model shifts (metered billing), or wailing to the heavens about the dire menace that is P2P traffic. "There is not a single sign of an unmanageable flood of traffic," Odlyzko says. "If anything, a slowdown is visible more than a speedup
," he says.
Odlyzko's findings are further supported this week by new research
e-mailed to us by TeleGeography, who annually monitor Internet backbone operators and track peak and average network utilization levels. Their findings? Things are just fine. "Despite predictions from some quarters that Internet traffic would overwhelm networks, the global Internet is far from reaching its maximum capacity," says the group.
That's certainly not what we were told in countless editorials like this 2007 one
from an AT&T-backed think tanker that hinted the end of the Internet was perilously close, made potentially closer by pesky network neutrality advocates. In fact, according to Telegeography, global backbone capacity has grown faster than Internet traffic for two years running. While strong traffic growth between the US and Latin America did outstrip supply, the majority of global links are actually seeing reduced strain, because, as you'd expect, core carriers have easily adapted.
For the second consecutive year, the rate of underlying international Internet capacity deployment outpaced global Internet traffic growth, leading to lower utilization levels on many Internet backbones (see Figure: International Internet Traffic and Bandwidth Growth, 2005-2008). Between 2007 and 2008, average traffic utilization levels decreased from 31 percent to 29 percent while peak utilization fell from 44 percent to 43 percent.
So if backbone capacity is fine and core carriers are easily keeping pace, what's the problem? No ISP would ever admit their last mile networks aren't the top of the line, but problems are frequently caused locally, where carriers aren't willing to take the investor knocks
necessary to upgrade capacity. It's generally a direct result of skimping out on upgrades to beat quarterly estimates, but it makes much more business sense to blame the problem on terrifying, nebulous and uncontrollable externalities.
The investment community
, incumbent policy groups
, lobbyists and executives are best served by the public believing that bandwidth is not just rare -- but an extremely endangered resource requiring new pricing models
, greater restrictions on usage
, anti-consumer legislation and higher prices. The term "exaflood" was actually coined by the same think tank
that spawned the controversial "Intelligent Design" effort. You should be able to recognize a similar disdain for real science. Are you still buying into the exaflood? You shouldn't be.
Have faith in broadband evolution.