As I said, I'm almost sure the figure is based on per address (Flat 1, Tower Block, Long Road, Big City, AB12 3CD) or (Big Farm, Country Rd, Middle Nowhere, Little Village, ZY98 X76)

As far as availability goes, if there is a method of getting a fixed service at 24Mb/s or above it will be consider Superfast and included in the 90/95% figures.

What do you mean by the fibre infrastructure supporting the demand of 24Mb/s, do you mean at a routing level or locally... also remember are we talking all these 24Mb/s utilising the full 24Mb/s simultaneously because currently I don't think the UK has enough infrastructure in place to cope with 90% doing 5Mb/s simultaneously.

A block of 200 flats counts as 200 premises, pretty simply really.

"The important question is however if the fibre infrastructure can support a 24 mbps demand if 95% of properties were all connected?"

Why should it need to, consumer broadband is cheaper than leased lines because it is a contended service.

"I'd say a much more acceptable definition would be if 95% of households and businesses are connected to a 24 mbps link or can be connected within 30 days of order. "
Urm that is what it is already, ok the 30 day order thing is not explicitly stated, but with FTTC that is normally the case, apart from when a cabinet is in a line card capacity needs upgrading phase.

That too, more effort goes into tracking Openreach, but then they have the largest expansion programme at this time. The VM coverage is tracking their changes, but usually on a monthly basis and the same with KC.

The number of journalists I talk to, who are shocked to hear the coverage levels as in is a lot higher for an area than they thought, and you then have to explain what this means e.g. demographics of an area. A common one is that they think BDUK is about enabling coverage in 90% of rural premises.

For the record I am part of the 4% of Surrey that has fibre based available and is not superfast - must be a miracle I can work from home if the needs of home working are true.

I'd like to see G.fast implemented next.

The increased use of fibre optics, and the reducing of the copper wire should also lessen the interference on lines. This for me is one of the big, not so talked about, improvements with G.fast.

Also the speed boost is a good thing too!

So really there's the lie in the current availability figures:

the system doesn't provide the advertised capacity, even allowing asymmetric delivery. There's non advertised contention.

The 80% or whatever availability only works if circa 30% of households use the service. If the take-up figure was say 75% of households sign up the fast broadband equipment would need major overhaul in the exchanges and at cabinets.

In reality there's only one third of 80% delivery of service and that is subject to contention.

I would think there is plenty of capacity given they can provide IPTV

In reply to a post by oaksys:

So really there's the lie in the current availability figures:

the system doesn't provide the advertised capacity, even allowing asymmetric delivery. There's non advertised contention.

The 80% or whatever availability only works if circa 30% of households use the service. If the take-up figure was say 75% of households sign up the fast broadband equipment would need major overhaul in the exchanges and at cabinets.

The availability figures are fine, it indicates that Superfast 24Mb/s> is available to xx% of premises, contention is advertised the way it always has been with the "up to" quote in advertisements.

As the new networks are built, more capacity is added to exchanges and cabinets are designed to serve a set number of connections and to manage the bandwidth passed through them,

cabinets like mine also get further capacity upgrades as take up grows, and what I can 100% surely tell you is I've never suffered any slow downs and congestion like issues related to the fibre cabinets or network locally even when the capacity "apparently" got low!

Just like the electricity grid, where if we all switch on our electric fires and kettles at the same time it would break the national grid.

There are commit rates over the cabinet back haul of 15 Mbps and 30 Mbps, and at the exchange it is down to the major providers how they contend the service and it is this peak time performance that can differentiate the services.

We run speed test analysis so can see how things change over time, and Ofcom spends a chunk of money monitoring it all too.

Is it correct that each card in NGA is 1Gbit yes?

If so a Huawei 288 (6 cards) 6Gb/s / 288 = 20.8Mb/s or similar or am I barking up the wrong tree?

I know multiple GigE is possible, but whether its delineated per line card, or for the whole DSLAM have never bothered to find out.