1 Comment

Summary:

So what’s driving investment and change to the core of the internet? Data is still important, but content is the change agent these days as content providers seek new paths to deliver an experience faster and closer to your home network.

Craig Labovitz DeepField Structure:Europe 2013
photo: Anna Gordon/GigaOM

A fundamental shift from data to content has actually changed the infrastructure of the internet–and the change isn’t over yet. That’s the view of Craig Labovitz, cofounder of DeepField, speaking on Wednesday at GigaOM’s Structure:Europe conference. And if anyone should know about the “plumbing” of the internet’s core, it’s Labovitz; he’s studied it for more than 15 years.

Through 2008, Labovitz says, the “core” of the internet meant 10 to 15 companies handling the long haul backbones. They incurred most of the cost for the core at that time, but provided most of the value and delivered the majority of internet data. That started to change in 2006, however, thanks to an explosion of content. “By 2010 Google went from 1 percent of traffic to 6 percent due to the YouTube purchase,” he noted, requiring changes to the core: Content shifted to data centers, which required massive investments.

What was the impact? In 2009, about 150 companies accounted for 50 percent or more of internet traffic. By 2013, just 50 companies deliver more than half of all internet traffic, and they do so using a smaller number of data centers and Content Delivery Networks. There’s now an incredible density of data coming and going through the core. Need an example? Netflix is it, said Labovitz, accounting for 35 percent (and growing) of U.S. web traffic.

Sure there’s plenty of data still racing through the pipes of the internet’s plumbing. But a specific type of data — content — is what’s changing how those pipes are arranged as content providers are looking to get a more localized presence closer to your home network.

Check out the rest of our Structure:Europe 2013 coverage here, and a video embed of the session follows below:

  1. What happens when we go to 30 4k (or heavens forbid 8k) cameras in a stadium and consumers have choice as to their preferred camera angles. One could easily imagine 4 x 500 meg (48 gig compressed) coming into one endpoint. Suddenly Google fiber looks paltry. Policy makers, pundits, trade management and investors need to wake up to the fact that this tomorrow is just 5-10 years away. Our current vertical service provider models are not going to deliver the solution.

    Share

Comments have been disabled for this post