Wednesday, February 14, 2007

Google Warns

New Internet TV services such as Joost and YouTube may bring the global network to its knees, at least one Google executive has said publicly.

At some level, of course, statements that "the Web infrastructure, and even Google's doesn't scale,” as Vincent Dureau, Google's head of TV technology says, are simple observations about the best effort nature of the Internet. It wasn’t designed for real time services. Unlike the global telecom networks, the Internet wasn’t built expressly incorporating quality of service measures.

At some level, Google saying today’s Internet is “not going to offer the quality of service that consumers expect” is simply an observation about the way the Internet now works, unaided by priority-enhancing measures or further raw bandwidth upgrades.

Google in Europe appears to be offering to work together with cable operators to combine video search and tailored advertising with the cable network access networks, which are, by design, closed and quality controlled.

That should have come as quite a comforting perspective for cable operators who have reason to feat the emergence of over the top video. At while there is great truth in what Dureau says, it may not, strictly speaking be completely true of all IP networks. Private networks can be created to ensure quality video delivery, even if will be difficult to ensure that the public network supports video using “best effort” delivery.

Compared to a voice-centric network, a video-capable network must support sustained throughput two orders of magnitude or 1,000 times greater than required for voice. That is not to say the Internet will not support video. It just might not be completely satisfactory as a means of delivering real time video. Downloaded video won’t be as big an issue.

In fact, Gartner Group analysts estimate that 60 pecent of the Internet traffic that is uploaded from computers is peer-to-peer traffic, mostly from consumers swapping films and TV shows.

No comments:

Many Winners and Losers from Generative AI

Perhaps there is no contradiction between low historical total factor annual productivity gains and high expected generative artificial inte...