IP Traffic To 'Double' Every Two Years

Stony Stevenson writes "Web traffic volumes will almost double every two years from 2007 to 2012, driven by video and web 2.0 applications, according to a report from Cisco Systems. Cisco's Visual Networking Index (PDF) predicts that visual networking will account for 90 percent of the traffic coursing through the world's IP networks by 2012. The upward trend is not only driven by consumer demand for YouTube clips and IPTV, according to the report, as business use of video conferencing will grow at 35 percent CAGR over the same period." I left the apostrophes around the word "double" in the title because the linked site has them, but for the life of me I can't figure out why.

'double'

[Swizec]

How do they even define what a double of IP traffic is? Double the amount of packets? Double the amount of connections? Double the amount of IP's used in traversal from one point to another?

Self serving?

[stokessd]

Hmmm, a huge supplier of networking gear is saying that network will continue to grow...This article is making me want to buy a lot of networking gear to get ahead of the impending doubling. I wonder if that's the intent? Nah, couldn't be.

Sheldon

Take advantage of IP Multicasting

[Twide]

To combat the increase of traffic why cant businesses/ISP's work more towards multicasting? for at least some part of the webs streaming sources (internet radio.. etc) this would alleviate some of the load surely..

Re:Self serving?

[Anonymous Coward]

"Hmmm, a huge supplier of networking gear is saying that network will continue to grow.."

I can definitely believe it, over the past year my ISP has been implementing very restrictive banwidth caps.

"Everyone's going to buy lots more routers..."

[Angostura]

.... router company tells shareholders.

Re:'double'

[postbigbang]

Although I'll not argue the data, it seems a bit self-serving to make a prediction like this. There are argument offsets to this data that might be salient:

1) video, especially HDTV, is being delivered by cable systems out-of-band of the Internet because of its high data rate. This trend will continue, else cable companies will have to evolve (and rapidly) immensely fast infrastructure that must also match CPE. Unlikely to occur. However, DSL providers are faced with a similar problem. What this means is that HDTV will be switched at the head-end eventually, and not 100% available to CPE. Video on Demand will become the rule of the day, thus offsetting some of the perceived growth in Cisco's numbers

2) business video conferencing, even in the face of $4 or $8/gal costs, just hasn't taken off. Codecs are available that can do a very good job of offsetting bandwidth needs.

3) isochronous media is still a bear, but it simply needs priority and priority in the face of network neutrality calls will be difficult without increasing bandwidth and therefore asset costs, which pays/plays into Cisco's hands mightily (are you watching, Wall Street?).

4) Cellular/mobile growth will climb, but it's more linear in growth and devices that receive entertainment content that uses bandwidth are largely distributed on private, rather than the public Interent. You just can't make a mobile phone in to an HDTV no matter how much you try, and the demand for it isn't there despite the best hopes of the telcos.

5) regionalization of content distribution is already occuring, and so a distributed infrastructure will 'cellularize' a lot of transfers. Transasction-focused systems aren't well managed through regionalization, but because entertainment systems aren't usually transaction-based, the use case is largely moot.

Doubling is therefore a projection based on a lot of assumptions, mostly favoring the maker of the study.

Re:Duh

[Anonymous Coward]

It will not take to many years of doubling before the amount reaches good enough.

Yeah because a real-time 3D virtual world interface run over our data pipes won't ever emerge or become standard; Or virtual machines usable from any terminal that move location in the grid; Or any other of a hundred things my little lizard brain can't conceive of yet.

How can people continuously make the statement that at point x we'll have enough that we'll stop expanding our data/memory/network capacity? How many times do we have to make that mistake before we realize the assertion is silly.