There are many numbers in this eWeek story describing the latest edition of Cisco's Visual Networking Index. My favorite is that by 2013, the company predicts that there will be 11 billion square feet of screen space on earth.
The entire story is interesting. The company says that by that year IP traffic will be five times what it was in 2008, and that 90 percent of that traffic will be video. Other nice tidbits include the fact that the number of exabytes of IP traffic will grow from 8 per month last year to 56 per month in 2013. Global IP traffic in 2013 will be 667 exabytes which, of course, is two-thirds of a zettabyte. A zettabyte, as we all know, is a trillion gigabytes.
So the bottom line is that IP traffic will proliferate at an astonishing rate, and that video will be the driver. There is absolutely no reason to doubt that this will happen, but there are interesting qualifications and questions that should be posed as well. The first, of course, is to keep in mind that the sponsors of the report are the preeminent networking firm in the world. So it's a bit like Coca-Cola predicting that in 2013 people will be thirsty.
The question of sponsorship always is relevant when a survey or study is done by a vendor who would benefit by a certain outcome. In this case, however, the Cisco presence makes it reasonable to assume that the study was objective. People indeed will be thirsty in 2013, and people also will need a ton of bandwidth. If Cisco says it's two-thirds of a zettabyte, that's good enough for me.
A second question is how meaningful predictions can be made in an atmosphere in which things are changing so precipitously. The answer, no doubt, is that they can't be. For instance, the advent of demand from smart grid and other green initiatives were largely unforeseen a few years ago. Now it is considered to be a major driver. Similar shifts almost certainly will occur during the term of the study -- and, perhaps, not all in the direction of more bandwidth. Despite the uncertainty, however, the takeaway that there is going to be a mind boggling and constantly increasing amount of IP traffic in general and video in particular pinging around the globe no doubt is solid.
The vital question, and one where the answer needs a bit of specificity, is this: Assuming growth of this scale-but also assuming the normal growth in available bandwidth, the improved capacity of the networking gear (by, for instance, opening new wavelengths in optical transmission equipment) and more efficient compression and related techniques-will all that demand lead to a crisis? That's a policy question for the government and a planning question for carriers that must be addressed as we approach the second half of 2009.
The question of how to prepare for the near- and mid-term future is vital. It's also apparent in the staggering move toward IPv6. The core of the issue is whether we are at the height of the growth curve in demand or if it will continue to accelerate. With apologies to Everett Dirksen, it is possible to say that exabyte here and an exabyte there pretty soon adds up to some real bandwidth.