Growing Demand for Bandwidth: A Rising Tide or Cresting Wave?

Andrew Odlyzko is the leading researcher of demand for bandwidth on the Internet. While some have been talking about explosive growth, he's quietly and continually shown through his research that while at one point the Internet was doubling, that at this point it's growth is pegged closer to 50%.

I just came across another article about him and his presentation at a recent conference where he shared that in a place like Hong Kong, where a lot of people are subscribed to ultrafast broadband connections, that while users consume six times the bandwidth as we do in the US, that their growth rate is less than 20%.

That's an interesting thought to consider as it suggests that the presence of big pipes alone will not drive exponential growth in demand for bandwidth.

One question it raises is whether or not these lower growth rates are upper end limits or merely temporary plateaus.

Of course, some of this has to be attributed to the fact that the higher the overall demand the harder it is to realize a high percentage growth rate.

But even still I find the position we're in kind of odd. There's no denying that demand for bandwidth is growing. Yet Odlyzko's studies show that perhaps it's not expanding as fast as we first thought.

But then there's the reality that there are a host of high bandwidth applications on the horizon. Plus the fact that most Internet users today don't use high bandwidth applications. So the possibilities are there to ignite another round of exponential growth.

Not only that, but that growth in demand seems almost inevitable given the presence of ever-larger pipes and increasingly powerful applications.

For example, when a user goes from watching a YouTube video to watching a full-length episode of Lost in HD, that's not just a doubling in demand for bandwidth, it can be a tripling or quadrupling.

Or what happens when we start moving from small window videocalls to fullscreen? That again is an order of magnitude more bandwidth, say from 200Kbps to 1Mbps, and I'm not even talking yet about HD or uncompressed HD.

And the reality is we might end up having capacity run ahead of demand, so once more users start waking up to the possibilities of rich media Internet applications they may skip YouTube and go straight to HD; they may bypass small-screen videocalls for full-screen if for no other reason than they can.

Yet at the same time, a theme I consistently hear is that while getting the early adopters on board is easy, convincing the next group of people to join in the broadband revolution takes a lot more work. So even though we've got more people going online and getting broadband every day, equipping them to take advantage of high bandwidth applications won't be easy.

But yet again there's a potential for explosive growth as I'm a firm believer that if you can just sit down and show someone what's possible, a light bulb will go on and they'll start transitioning into the mindset of a power user.

So here we are, in a position where the reality doesn't match up to the promise and an uncertain future lies ahead. What would happen if a bunch of celebrities or politicians started pushing the benefits of video-on-demand online and videocalls? It's quite easy to imagine a tidal wave of interest crashing onto the Internet.

The reality is that any upsurge in demand can take down individual servers quite easily and could even threaten large swathes of networks as everyone who knows about this will tell you that the Internet can't yet support TV-sized audiences.

But at the same time, all trends point to modest growth that will gradually increase over time.

The scary thing is that ultimately we have no control over this. Part of me wants to encourage this huge upswell in interest as that's ultimately one of my big goals, to get everyone engaged with using the Internet to better their lives. But another part of me worries about what might happen if I'm successful in that goal as not only are the networks today not capable of handling unlimited demand but the long-term trends in demand that are driving investment don't suggest the near-term need for all that much more capacity than we have today, making it less likely the networks will be there to support a significant uptake in demand should it materialize.

What troubles me most about all this is that this issue totally frames the decisions being made in how much to increase capacity, and yet there seems to be a giant gap between the reality and the potential, and what's so uncertain is how quickly we're going to reach that potential. Will it be slow and steady growth from here on out? Or will something happen to ignite a firestorm of demand?

Fundamentally I'm a big believer that the demand for bandwidth is there it's just bottled up in the fact that most people don't understand today what broadband can mean for their lives. And I do believe we can realize a future soon where it's not just the early adopters using high bandwidth applications.

But the most worrisome part is we really don't know what the future will hold. Some predict even less than 50% growth, but what if the opposite is true?

What happens when someone decides to put the Super Bowl online and everyone shows up at the same time?

What happens during the next big national emergency and everyone's going online to find out more information?

Will the Internet be able to support these surges in demand? The truth of the matter is no one knows.

And unfortunately the only way we're likely to find out is to have a surge that overwhelms the Internet, leaving us without its power potentially during a key time.

I wish there were a way around this but I'm not sure there is. But I do know that unless we can decide where we're at, either riding a rising tide or poised on the precipice of a crashing wave, we can't properly prepare to survive the aftermath.

And it's part of my life's work to prove that the wave is not only cresting but it's ready to crash in many splendid directions as we all start to realize the full power of broadband in our lives.