There has been a lot of speculation in recent weeks about whether global oil demand could soon outstrip global production capacity. During a panel discussion sponsored by the Internet Innovation Alliance, several experts pondered whether something similar could soon happen with bandwidth.
The panel, dubbed "The Cost of the Exaflood," addressed problems that ISPs could experience as they try to keep up with the increased global bandwidth demand currently being driven by high-bandwidth applications such as peer-to-peer transfers and HD video streaming. Craig Moffett, a senior analyst for US cable and satellite broadcasting for Sanford Bernstein, made the gloomiest predictions by saying that it was inevitable that major ISPs and content providers would experience significant losses in revenues as a result of increased bandwidth demand.
"We've got all kinds of stresses on the system right now, and I suspect that... we're going to see some pretty significant blow ups over the next few years," he said. "The rhetoric around this doesn't remotely match the economics."
Gary Smith, CEO of network infrastructure vendor Ciena, agreed with many of Moffett's points and said carriers would have to consider new pricing models that differ from the current model in which users pay a flat rate for a certain amount of bandwidth. Instead, he recommended that carriers begin charging users rates in direct proportion to the amount of bandwidth they consume.
"We're seeing a tremendous expansion of applications," he said. "The issue is, who pays for it and over what period of time are people going to get returns from that? Part of that responsibility goes to end users... We're going to have to get to a point where people pay for usage in some way. You can't continue to have a system where someone who uses just a small amount of bandwidth pays the same amount as someone who is streaming video for 24 hours a day."
Panelist Johna Till Johnson, CEO of Nemertes Research, agreed with the other panelists that ISPs would face some major challenges in making sufficient revenues as they upgraded their networks to handle increased bandwidth demand. The big problem, she said, was that while Web traffic has been growing at an exponential rate, the access capacity has been growing linearly. Thus, unless something changes in either traffic demand or access supply, Johnson predicts that there will not be enough bandwidth to meet global demand.
But unlike Smith or Moffett, Johnson did not explicitly endorse a tiered system of charging consumers rates that are directly proportional with the bandwidth they consume. Instead, she said no one currently has enough concrete knowledge about traffic growth and access capacity, and recommended that the government encourage ISPs to more openly disclose their traffic data so that researchers could get an accurate picture of how much bandwidth demand is growing.
Beyond this, Johnson said any national broadband policy to help mitigate the bandwidth crunch would have to consider trade-offs between ubiquitous access and innovation. For instance, if the government were to pay for access to rural areas that ISPs can't provide because of revenue constraints, then she said those connections would very likely become inferior to other connections within a matter of years.
"If your overall goal is ubiquitous connectivity, you're basically doing that at the expense of some level of innovation," she said. "For instance, consider what happened after the government took over the subway system after private subway companies went bankrupt during the Great Depression. The good news is we still have a pretty good subway system today. The bad news is I don't think we've actually changed and updated and innovated much about it for years."
Nemertes Research first started sounding alarm bells about the potential pitfalls of exponentially increasing bandwidth demand last year when it released a study claiming that the Internet could overload and lead to brown-outs in two years unless backbone providers invest billions of dollars in new infrastructure. Net neutrality advocates have typically derided concerns about an upcoming "exaflood" as exaggerated, and point to research conducted by the Minnesota Internet Traffic Studies project showing that while overall traffic has grown year over year, the actual rate of traffic growth has slowed in recent years.