Who funded those studies?
I would like to mention one thing that came to mind immediately when I read this, as there was a similar report from the UK just lately...
If you were a caterpillar happily chomping on a lettuce leaf, and someone told you that you could change into that beautiful butterfly in the sky, would you believe them? The same thing applies with a lot of people, they can't yet imagine the future, but it will come, and we have to be ready.
The current infrastructure cannot be tweaked forever, we have to invest in what will deliver gigabits one day. It matters not that 10 meg is probably adequate at the moment for a lot of people. The key thing in all this is not to waste public money on something that isn't futureproof. If we are going to build something, lets build it once, and build it right. If its over capacity, so what? It doesn't cost more to lay fibre than copper, in fact its probably cheaper. Fibre networks are cheaper to operate and maintain.

There should be an abundance model, not a scarcity one. Diamonds and gold are only precious because they are scarce. If the streets were paved with them we wouldn't value them. The same with the internet. Because it trickles down to us through narrow pipes it costs us too much. Put in the fat pipes and get them everywhere, it creates a golden path to the future. Lets have an abundance. The countries who invest now will reap the rewards in the future.

The fact that high-speed broadband has not been largely deployed is likely to have constrained the development of bandwidth-hungry applications – a chicken-and-egg conundrum probably exists. It is possible that wider utilisation of bandwidth-hungry high-definition telepresence-type applications could reach the mass market if sufficient users possessed the required bandwidth. The recent introduction of videoconferencing-style facilities for Facebook and Google+, requiring up to 8Mbit/s, is perhaps an example of the way developers are utilising the increasing bandwidth provided to many users.

However, this observation is in line with some who try to justify high speed bandwidth by suggesting there will be a ‘killer app”. I disagree with this view. I also disagree with many studies that extol the virtues of high speed broadband without considering whether the applications that they suggest are so useful really require increased bandwidth. Studies hyping the need for bandwidth for enhanced deployment of smart buildings and grids are prime (misguided) examples.

While there will not be a killer app, it is already likely that, in aggregate, many of the bandwidth consuming applications used in a modern home will necessitate more bandwidth. For example a household streaming two TV channels in high definition (requiring perhaps 20–25Mbit/s in total), playing high-definition online games (requiring up to14Mbit/s), while taking part in multiple-user Google+ online conversations (requiring 8Mbit/s) would need 40–50Mbit/s in total. Businesses using four good-resolution CCTV cameras (requiring 5–6Mbit/s in total streaming to a remote location), with six staff connected to cloud-computing applications (requiring, say, 12Mbit/s) with a high-definition telepresence application (requiring 24Mbit/s) would need a broadband connection with speed of more than 40Mbit/s.

It seems reasonable to assume that bandwidth-hungry applications will develop more rapidly in the future if the number of potential users of this type of application increases, since this will make the economic case more viable for developers.

Agree Paul. Add on to that the granny living with a family and her telemedice device ceases to function because the grandchildren are hogging the bandwidth...
... the killer app is active families, all using the internet for their own needs at the same time. Or the busy business.
Add to that the throttling and capping needed because ISPs are stuck buying off a monopoly who is moving investment into content instead of infrastructure, and who pays its leaders £5 million bonuses and you can see why we have a fight on our hands.
People always want the cheapest, and having got it they then moan its too slow. Leased lines have always been available and still are for serious business use, but outpriced for families and unavailable for rurals. We have to break free from the copper mindset protecting the scarcity model and get the fibre roads built. Then whatever we need or want will be available at an affordable price.

Today seems to be my day for agreeing with everything. Paul is quite right (Hi Paul!) there will be no killer ap and it is a mistake to look for one. Instead we should look at the aggregated use of bandwidth.

As for Tore's view about excess capacity I believe that the question is the wrong one. ( I will read the detail of the report however - promise) Ask what you can use the network for and you will come up with an answer that is dictated by the capacity of the network which is not the same question as what could you use the network for? There is excess capacity in the fibre networks in the UK because laying fibre capacity is very cheap once the hole is in the ground. However, the fibre doesn't go to the places where it can meet the demand side.

I keep on referring back to the Plum report from 2008 on the "Framework for Evaluating the Value of Next Generation Broadband" it supports the view taken by Paul that the aggregated demand of business, residential services and the requirements for new content services are the areas where the value lies and therefore the need for capacity.

At the moment we seem to be applying the value proposition to a scarcity model (a bit like gold) but the evidence suggests that the value is in the services which, in turn, create the demand.

I am looking forward to discussing this further at DA12.

Since when did playing hi-def video games require 14 Mbps.

The state of the art FPS, use perhaps 100 to 200 Kilo bits per second, even when voice is included.

Services like OnLive which use cloud computing to host the game and serve just VIDEO, would use 14 Meg, but will never approach the 1.6Gbps uncompressed rate needed to look the same as a locally rendered game.

Fibre networks, are not just about raw Gbps, but about allowing low latency network, and supporting multiple services at once. i.e. 1 x HD stream in lounge, music in teenagers room, another person gaming, while social network updates wander in.

Far too many people have 5 to 10 Meg type connections now, and because gaming is poor assume they need a faster link. When its just provider is congested and thus poor latency performance.

Your point is well made Andrew and thank you for pointing out that we are omitting the valid point about latency.

When all three of us are at home I have counted: up to four laptops, one iPad, two iPhones, one blackberry, two games machines, one ip enabled blue ray player, one femto cell.

Apart from background tasks such as e-mail a check on our ip stream shows two facebook accounts, one google plus account, two Twitter accounts, one on line game in progress for most of the evening, face time, skype, i-Tunes, i player and let us not forget internet browsing which since the advent of the iPad is most of the time.

We only average 12 Gigabits per week of downoad but we are unable to use most of the services available via XBox or Sony because of latency issues and we accept the quality issues with Skype and Google Hang outs. This figure includes work which is a lot of on line research.

My point is that there is pent up demand in one household and to read the list you would think we did nothing else which is not the case, quite the contrary.There is an assumption abroad that people use the Internet sequentially viz if you're doing one thing, you can't possible be doing anything else which is not the case. We need to start to consider the question of ubiquity and pervasiveness as well.

12 gigabits a week says it all... on 3g dongles the cost for that is around £60. Impossible with satellites, and my daughter has a 10 gig a month cap on her account. You are spot on with the scarcity model! It also goes to prove 4g probably won't work even in areas its available as that will probably cost the same as 3g.

The cost of 3G/4G will be mainly due to the scarcity of capacity on the wireless spectrum, rather than the cost of the fibre backhaul.

Sky is already putting 10GigE links into some exchanges for its LLU backhaul. With 10GigE the cost of the hardware to go on the ends of the fibre gets to be significant. Even in data centres firms who have fibre already in place, have to plan for the upgrades in hardware.

Love the origin of this post - you read my mind. One could get the idea that once a high speed is provided by one of Europe’s many market players, it has to stay at that speed forever. But we’re talking about a dynamic sector here, right? I think Europe needs to give itself less of a hard time and be more optimistic about its ability to roll with the punches and adapt speeds according to market demand. It’s too easy to blame policy makers for being prescriptive and we should definitely have some lofty goals in mind. This exchange touches on demand – i just wish public policy could focus on the demand element, too!

Once a high speed is provided??? what do we define as a high speed, that is the question really. If you give a rural person 2meg they would think it was great after dial up. Do you remember throwing your old dial up modem on the scrap heap?
If you give someone on adsl close to a cabinet a fttc connection they will also probably notice improvement.
We can't do this though, because once we invest in cabinets they become the noose round our necks. We have to think to the future, and a little fibre laid now can serve us well for decades. All that has to change is the lights on the end. The fibre itself is limitless, unlike the old phone lines. If building fibre networks are what you call lofty goals then I totally agree Caroline.