Navigating the Exaflood

On June 1st Cisco launched the fifth edition of the Visual Networking Index (VNI), an excellent research material that forecasts global IP traffic growth for the years 2010-2015.  The report confirms that we are entering the Zettabyte era, as the total amount of global IP traffic will quadruple by 2015 to reach 966 exabytes per year.  As difficult as it is to imagine such an order of magnitude, just think of the equivalent of 28 million DVDs being downloaded every hour. That’s the volume of traffic that global IP networks will be supporting in 2015.   For those readers with an affinity for history or literature we can also use the very graphic “LOC pattern,” which illustrates that 966 exabytes is the digital information equivalent to the content of 50 million Libraries of Congress.  Perhaps in the future only Borges’ Library of Babel will be an adequate frame of reference…

The news may come as a striking surprise for some people outside of our sector.  However, it was over four years ago that Bret Swanson, working at the Discovery Institute at the time, and later at the Progress and Freedom Foundation, used the term “Exaflood” to describe the growing wave of broadband Internet traffic.  Giving credit where credit is due, back in April 2008 at the Westminster e-Forum on Web 2.0 in London I happened to hear Jim Cicconi, AT&T Senior Executive Vice President for External and Legislative Affairs, warning that by 2015, Internet traffic would be fifty times greater than it was in 2006 and that we would be looking at a traffic load of 1,000 Exabytes.

My only concern with the forecasts of Cisco’s VNI is that when analyzing the driving forces behind this dramatic traffic growth, faster broadband speeds are mentioned together with the increasing number of connected devices, Internet users, and the proliferation of video.  In fact, there is an assumption that global broadband speed will increase fourfold between 2010 and 2015, with the average broadband connection moving from 7 Mbps to 28 Mbps download capacity.  In its analysis of the data for Western Europe, Cisco predicts that the average broadband speed will grow from 9.2 Mbps to 36 Mbps in 2015, surpassing the US, which will move from 7.5 Mbps to 27 Mbps.  That prediction seems to be overoptimistic if we consider the broadband targets set by Commissioner Kroes in the Digital Agenda for Europe, where the goal is to achieve broadband coverage at 30 Mbps or more for 100% European citizens in 2020.  And it seems difficult to match Cisco’s prediction with the latest Commission Digital Agenda Scoreboard, which indicates that in 2010 only 5% of all fixed lines in the EU were delivering speeds of 30 Mps or more.

The bold presumption of a fourfold increase in broadband speed runs the risk of taking investment in broadband capacity for granted.  And as we know from the AT Kearney Report on the Economic Sustainability of the Internet, large-scale investment in fiber or LTE will only happen if the right conditions to navigate the Exaflood challenge are met. This calls for all stakeholders in the ICT ecosystem to share responsibility for the sustainability of the Internet by promoting a new model based on commercial flexibility. We will need to consider exploring new approaches to interconnection, including differentiated offers at the retail level and at the wholesale level through end-to-end guaranteed QoS, as well as managed services in access networks to tackle the challenge of exploding video traffic on networks. Policy-maker on national and international level as well as Regulators should focus their attention on these issues if they do want to achieve the desired levels of high-speed network infrastructure in their territories.

What are your views? Do you foresee any alternative business models, regulatory conditions or technology developments to navigate the Exafllood? What should be done by policy-maker to tackle that challenge and at the same time foster broadband roll-outs?