UPC go to 200Mb/s; digital divide worse than ever.

clohamon wrote: »

Eircom originally planned to do FTTH rather than FTTC, and for all the right reasons, i.e. infinite headroom, lower maintenance, single-step upgrade, no premature FTTC asset write-offs etc.

The financial pressure and the speed at which UPC were consuming their urban market probably explains why they went for the quick, cheap fix, i.e. FTTC.

You're right, its looking like a bad decision, but it was probably the only decision they could make.

I'm not sure it was a bad decision.

I agree that I originally thought that going for FTTC over FTTH when they faced such strong competition from UPC was madness, but now having seen the speed and quality of Eircoms FTTC rollout, I'm actually pretty impressed.

The reality is for the 500 million Eircom are spending, they would have only been able to cover a quarter of the homes that they can with FTTC and even at that a FTTH rollout would have taken much, much longer.

The reality is that the FTTC rollout is bringing pretty decent level of BB (70mb and 100mb soon with vectoring) to the majority of homes in Ireland, including many not serviced by UPC, at a very fast pace.

To be honest, 50mb/s is more then fast enough for the majority of people for the foreseeable future *

I also wouldn't say that Eircoms FTTC isn't competitive with UPC. Eircom offer upload speeds of 20mb/s, while UPC's max speed is 10mb/s

In the age of youtube uploads, dropbox, etc. I'd actually take 70mb/20mb over 120mb/10mb. At speeds above 50mb/s you are usually been throttled by the download servers anyway.

Also I wouldn't say that the FTTC assets will be prematurely written off. From what I've heard the whole FTTC network has been designed with FTTH in mind.

The FTTC cabs fully support FTTH too and seemingly Eircom have run enough fibre drops to each cab to support FTTH from the cabs in future. All work that would need to be done if you went FTTH anyway, so no real lose there. The only lose you have is the engineer installs currently being done by for FTTC and potentially if they had went FTTH from the start, they might have been able to use cheaper FTTH only cabs.

But then you have to balance that with a much slower and more expensive rollout. I think Eircom have made the right decision.

* I know saying this sounds crazy, I remember 10 years ago being at Eircom and IOffL conferences and laughing at people saying that there was no demand or need for broadband. So I short of cringe when I hear myself saying this. But realistically I don't see any applications coming in the short to medium term that require speeds much faster then 50mb/s

50mb/s is plenty fast for a couple of HD video streams. The only technology that I can see coming that might need more bandwidth is 4k video, but I'm not convinced that 4k video will take off in the same way as HD did. The reality is that you need a 80"+ TV to take advantage of 4k video and that is a size far too large for the majority of Irish homes.

Sure some totally unimagined technology might come along that requires gigabit speed internet, maybe something sci-fi like being able to plug into and share someone elses experiences, but nothing that we currently know about coming up is going to need anything faster then 70mb/s

bealtine wrote: »

I concur with most of this as it is indeed what is happening on the ground but I'm not sure if I agree totally. Remember the great pronouncements of the past like no computer will ever need more than 512k memory (MS) or there will only ever be one computer (IBM)? So I'm wary of saying things like "fast enough" etc, sure it will bring the speeds up and be great for urban dwellers but do we abandon the rest?

I appreciate what you are saying and agree with you to a certain extent. However back 15 years ago, those of use working in the IT industry could see the upcoming revolution, things like online gaming, streaming music, streaming HD video, cloud services, video calling, etc. They were all in their early stages at the time and we could all see the need for high speed BB to deliver these services in the future. A future that has now come to pass.

The same isn't true for gigabit BB services. I just don't see any applications that will be coming in the next 10 years that require gigabit speeds.

I agree some unimaginable, sci-fi breakthrough might come, but I wouldn't go investing billions on infrastructure just in case something like that happens.

bealtine wrote: »

Perhaps the FTTC rollout could be seen as a first step towards what should be the ultimate goal of FTTH?

Of course and clearly Eircom has designed their FTTC rollout as such. Future proofing it for FTTH.

And I believe Eircom may start rolling out FTTH in future for a few reasons:

- Greatly reduced maintenance costs of fibre versus copper
- To compete in marketing terms on speeds with UPC and ESB fibre if it happens.

But I don't think they necessarily made the wrong decision with going FTTC first.

Eircom are investing 500m on this rollout, that can get you either:
- FTTC 1 million homes getting 100mb/s BB (when vectoring kicks in) very quickly (basically in pretty much one year).

or alternatively for the same money:
- FTTH 250,000 homes getting gigabit ethernet and probably taking 4 years to do.

I think the former is better for the majority of people. More people get a very good, usable speed sooner, rather then less people getting a fantastic speed that they don't really need.

clohamon wrote: »

What's actually re-useable inside the cabinet?

Well the cabinet itself, power and the fibre cables running to it.

The type of FTTH that Eircom is planning to use and has been trialling is GPON. See details here:

http://www.nextgenerationnetwork.ie/downloads/programme_overview/industry_overview.pdf

GPON requires kerbside local cabinets. These cabinets would contain optical splitters which take a single fibre cable from the exchange and split it to 32 fibre cables which then go to 32 individual homes.

Eircoms VDSL cabs support 192 phone lines, I believe Eircom has run a minimum of 6 fibres from the local exchange to each cabinet and each cabinet has space for 6 optical splitters, thus each cabinet is capable of replacing the 192 twisted pair lines with 192 fibre lines (6 optical splitters x 32 lines each).

So even if Eircom went straight to FTTH they would still have had to install the 5000 eFibre cabs and they still would have had to run 6 fibre cables to each of the cabs and all the civils related to that.

The only waste here when they do go FTTH are:
- Replace the DSLAMS with the optical splitters in the kerbside cabs
- Replace the NTU's with a ONT at the customers premises

Obviously they will also need to run fibre drops from the kerbside cabs to each home, but they would have had to do that under either scenario.

So there is a little bit of wastage, but far less then people are making out. The majority of the cost of rolling out either FTTC or FTTH is labour and civils. Replacing DSLAMS and NTU's is a relatively minor cost in the overall scheme of things for a project like this.

I estimate that the cost to replace the FTTC gear with FTTH gear would be about 50 million to 100 million (probably closer to 50). However the cost of doing the fibre drops from each cab to each home would be in the region of 1 billion or more!

It has been estimated that it would cost 1.5 billion to do FTTH to every home in Ireland. The 500 million invested in FTTC equates to about 450 to 400 million of this investment.

So you see, writing off 50 million for the FTTC gear when they go FTTH is relatively minor in the bigger scheme of things.

And the advantage is that they get to cover a large number of their customers, very quickly (much quicker then FTTH would allow) with reasonably fast BB, which helps stem the lose of customers and it also goes a long way towards FTTH, which they can then gradually rollout at a more consistent pace.

I believe going from 24 -> 70/100 -> 200+ quickly and cheaply is better then going from 24 -> 200+ over a much longer period of time and at an additional 1 billion cost!

Yup, I'd much prefer Eircom to focus on getting 50mb BB to every home in Ireland, then spend a billion getting FTTH 200mb+ to the same old urban dwellers, many if not most already have 200mb UPC available to them.

Thing is even if Eircom do spend a billion+ on FTTH, there is still no guarantee that it would be good enough to compete with UPC.

The FTTH service that Eircom trialled was only 150mb/s (which is a contention ratio of 2:1), which has already been surpassed by UPC's new 200mb/s service. UPC's current modems are capable of at least 400mb/s!

The coax cable UPC use is theoretically capable of 5Gb/s shared, note Eircoms GPON is capable of 2.5Gb/s shared *

UPC has proven itself to be very aggressive competing with anything Sky and Eircom do, so I'd expect UPC would match anything Eircom would do. And don't forget, UPC has their own fibre to the curb network, from which they could deploy FTTH for about the same investment cost as Eircom if they wished to do so.

So even if Eircom spent a billion+ doing FTTH, there is no guarantee it would help them win back customers from UPC.

* 5Gb/s is the max theoretical bandwidth of the cable they use, it would require them switching off their analogue and digital TV services and going all IPTV and it would also require new modems and "exchange" side gear, but it is possible.

Eircoms 2.5Gb/s GPON is shared amongst 32 users. UPC 5Gb/s cable is also shared, but harder to say amongst how many users as it depends on how tight their cable rings are.

Actually it turns out I was mistaken about the cost to do FTTH to every home in Ireland.

The estimated cost is actually 2.5 billion, not 1.5 billion:

http://siliconrepublic.com/comms/item/17057-tifx2010

I had read it as 1.5 billion in an Eircom presentation, but I guess that is just to do urban areas (60% of population), much of which is already covered by UPC. The 2.5 billion is probably needed to do every home in Ireland.

So that should be kept in mind when comparing the FTTC investment versus FTTH.

clohamon wrote: »

There's very little detail in that IBEC(TIF) document. Analysys Mason were extrapolating figures from their study of the UK carried out in 2008.

True, but the figures from this report are in line with my own figures that I estimated a few years ago. I estimated at least 2 billion for every home in Ireland.

Mason where estimating a cost of about €1,600 per home * 1.6 million homes gives you a figure of 2.5b+ €1,600 per home is a reasonable estimate in line with what other European telcos have found for their fibre rollouts. In fact it may even be on the low side.

clohamon wrote: »

Do the optical splitters need power?

No, but the cabs do as they are actively managed and have monitored alarm systems, etc.

BTW just came across another interesting report comparing the cost of FTTH to FTTC for both phone and cable companies:

http://www.analysysmason.com/Consulting/Client-Presentations/Regulatory-and-policy-development-in-NGA/Understanding-the-costs-of-NGA-networks-Ian-Struele/

Interesting article here on ultra HD TV's (4k and 8k) and their impact on broadband networks:

http://www.analysysmason.com/About-Us/News/Newsletter/FTTH-connected-TV-Apr2013/

Interestingly the article shows that FTTC networks should be able to handle 4k TV's (80" or greater) specially when bonding is used. The need for FTTH only kicks in when you go to 8k TV's (100" or greater).

As I pointed out the article also says:

Take-up of more-advanced UHD TV may be limited by non-technological constraints. There is a limit to the angular resolution that the human eye can process, beyond which there is no perceptible advantage to greater TV resolution. In order to make the increased detail visible, a larger screen is needed. The 4K sets currently available all have screen sizes of more than 80 inches; 8K TVs may be even bigger. Eventually the required screen size may become so great as to become an obstacle to take-up, whether because of the physical size of rooms, or the cost of super-sized TVs.