This report card in Popular Mechanics rates 5 IPTV services. Some of them look like they will be winners. As someone without cable or satellite TV, I should be grateful. But Blindside has spoiled my innocent enjoyment. What will be the impact on Internet performance? In June of this year, 2.5 billion videos were downloaded from YouTube, according to Reuters.

If, as I think inevitable, the UK government migrates essential citizen services online, can commercial offerings degrade Internet performance to such an extent that it threatens those services? Could telecare, remote surgery and e-999 services be crowded out by new networks hogging the Internet? Well, no–it might even help it. Pure IPTV doesn’t really use the Internet–it’s more like a different cable channel into the home. But it is stimulating development of the technology to get it from the narrowcaster to the PC or TV, and some of that technology and associated infrastructure might actually improve the Internet experience. But in parallel with pure IPTV, video and movies are being delivered over the Internet, and that traffic is growing rapidly. Is it growing quickly enough to justify concern?

In China right now, there are 100 websites trying to do what YouTube does. Telegeography estimates that international Internet traffic grew by 57% last year (down from 74% the year before). However, the firm estimates that capacity grew by 68% last year… so is there a problem? Here the question becomes more business than tech. I suspect there is latent demand for video and movies downloaded over the Internet that will respond very quickly to an increase in supply, especially if supply comes with new technology that improves the experience. The growth in IPTV may provide this indirectly.

As HDTV becomes the norm, each video will be between 4 and 8 times as ‘heavy’ as a normal video today… and although there is unused capacity in the system today, if Internet consumption of rich media explodes (as I think quite likely) that unused capacity will get sucked into service quite quickly.

A company called Black Arrow has estimated that in the US alone, viewers will watch 8 billion hours of online video/TV/movies in 2008 (compared to 376 billion hours of ‘linear’ TV consumption–don’t these people do anything else?). And as you might expect, Cisco is doing its own research, (scroll down halfway) which estimates that total Internet traffic will reach 18 exabytes (quintillion bites) per month, of which 11 of those quintillion will be the transport of commercial video on demand. The article linked to here estimates that necarrier Ethernet equipment spending by providers will double to £7.2 billion annually in the next 3 to 4 years.

I hope it’s enough. Because I didn’t even talk about P2P networks or the growth in traffic associated with mobile use of the Internet, let alone impacts associated with pervasive computing…