Imagining infinite bandwidth (and zero latency)

The idea behind the Infinite Bandwidth, Zero Latency strand at this year’s FutureEverything conference (Manchester, 12-14 May 2010 www.futureverything.org) is to try and get beyond the stale thinking about what people will be doing with much faster broadband speeds promised by ‘next generation’ access networks.

Next Generation Access or NGA is not precisely defined but is usually taken to mean internet access at very much higher speeds than ‘first generation’ broadband. That could mean access using wireless, ‘fibre to the premises’ (where each home and business has its own fibre connection), or enhancements of the existing telephone and cable TV networks. A tighter definition would demand a generational shift bringing more than just higher download speeds. For example symmetric bandwidth (same speed up and down), lower latency and jitter (the time it takes for a bit of data to travel from server computer to the user), and higher quality of service so that the speed doesn’t vary. By this definition, NGA really requires fibre to the premises (FTTP) or some of the more exotic varieties of wireless.

No demand

Only two years ago it was commonly said that there was no evidence of any demand for next generation broadband access and that as a result we here in the UK could afford to wait and watch before making any rash investment decisions. Usually what was really meant was that there was no evidence of a willingness to pay more for much higher download speeds. The problem was that it was the wrong question being asked. Had consumers and businesses been asked whether they would like to be able to send information much more quickly – to use a backup service for example, or upload a video so that the family in Australia can see it – without a big premium in cost, then the answer would doubtless have been different.

An even more obvious flaw with the argument was that it misunderstood the relationship between technological change and demand. In 1992 there was (of course) no evidence of any significant demand for access to the internet at all. The demand followed the supply. Luckily the investment required to start offering internet access services was low enough for small companies to do it anyway. Replacing the UK’s antiquated copper telephone and cable TV networks with fibre will cost a great deal more.

Things have moved on however: it is now at least accepted that NGA is needed now or very soon, and that the fibre variety of NGA will be needed eventually. But the thinking about what we will be doing with NGA hasn’t moved on. In particular, we are stuck with the notion of “superfast broadband”, the idea that it is all about how fast you can download things from the net.

Lessons from the past

Before dial-up access to the Internet became available in the early 90’s, some businesses and organizations were already using on-line information services – services such as news archives, scientific data on chemicals, flight information. There were the beginnings of consumer services as well – particularly in France – access to bank accounts, message boards and travel information. It was easy to see that computer networking would grow and that it would mean more of these services. But no one foresaw the world wide web, or the fact that within 15 years there would be a vast amount of information available on line for free and spread across millions (rather than thousands) of servers.

Similarly before the advent of first generation broadband at the turn of the millennium it was easy to see that this would allow new bandwidth hungry services like video. But no one foresaw YouTube. Even more importantly, by focusing on the higher download speeds that broadband would bring many people failed to realize that the really big impact would be that broadband is ‘always on’. This transformed the way that people used the net and led to the explosion in ‘web 2.0’ services like Wikipedia and Facebook. It also led to the explosion in file sharing that has completely destroyed some business models.

Now with NGA on the horizon, the thinking is similarly constrained. The focus is again on higher download speeds. These are important: you can do lots of new things if the data is arriving at 100 Mbps and more. But just as important are much higher upload speeds, higher quality of service that will mean that speeds don’t vary, and much lower latency.

To have understood how first generation broadband would change, well, everything really (we have hardly begun to understand the long term impact of Wikipedia – of having such a vast information resource available to 15% of the world’s population for free), we should have thought more about the implications of an always-on connection and upload speeds 5-10 times faster than dialup. These are what led to Web 2.0.

How to imagine

To understand how NGA will change everything again we need to think about the impact of much, much faster speeds – in both directions – and much, much lower latency. That means thinking beyond the usually-quoted applications for NGA such as ‘telepresence’ (ultra high quality video conferencing), cloud computing and telehealth. For example, how might cloud computing evolve if we were using the processing power of all the computers connected to the net, not just the servers? Rather than worrying about whether there is real demand for 100Mbps and more we need to start thinking about what becomes possible when bandwidth is no longer a limiting factor, when latency starts to approach zero (notwithstanding the limitations of the speed of light).

That’s why this strand is called ‘infinite bandwidth, zero latency’ – because it’s only when we think beyond what we can do with a lot more bandwidth and start thinking about what we would do if these constraints didn’t matter any more that we will begin to imagine.