CityFibre Start 10Gb XGS-PON Upgrade of UK Broadband Network
CityFibre has confirmed that they’ve started the national UK rollout programme for the XGS-PON technology upgrade to their existing GPON based full fibre (FTTP) broadband ISP network, which among other things will enable them to support symmetric broadband speeds of up to 10Gbps (9.953Gbps to be exact).
The work forms part of CityFibre’s wider effort to cover up to 8 million UK premises (funded by c.£2.4bn in equity and c.£4.9bn debt) – across over 285 cities, towns and villages (c.30% of the UK) – by the end of 2025 (here). The operator has so far covered a total of 2.6 million premises (March 2023), but only 2.3 million of those are currently considered to be Ready for Service (RFS) by an ISP.
However, until now, the vast majority of their network has been based off Calix’s GPON (Gigabit Passive Optical Network) technology and this invariably places constraints on how much capacity can actually be delivered to each end user (i.e. GPON shares capacity of 2.48Gbps downstream and 1.24Gbps upstream between 8 to 32 users).
Advertisement
The solution is to upgrade the hardware at the end of the fibre cable to kit that supports the modern XGS-PON standard, which CityFibre has chosen to do with support from Nokia and Calix. This will enable the operator to support more customers on each port of their Optical Line Terminals (OLT) and to keep pace with rivals that are already building XGS-PON, which in turn means the potential for symmetric speeds of up to 10Gbps.
As previously reported, CityFibre has already completed a pilot deployment of all this in the city of York, which confirmed various network cost savings, reduced power consumption, better performance and improved efficiency. The operator’s plan for their XGS-PON deployment was to work with their ISP partners in order to upgrade end-customer Optical Network Terminals (ONTs) – the modems that go on your wall – in phases.
Rollout Status
The original announcement, which dropped in July 2022 (here), said that the operator would start their main rollout from April 2023, but until now we’ve been unable to get official confirmation that this process had actually begun. The good news is that we’ve finally got some new details to share, although a full announcement is still expected to follow in the new future.
As it stands, York is already totally XGS-PON enabled, with only some customers still left to be migrated (this is managed by ISPs). But the wider rollout is said to be going fast, and they’ve already enabled 18 Fibre Exchanges (FEXs) across 13 of their UK towns and cities (for context – they’re currently live or building FTTP across 86 active locations).
Advertisement
The current expectation is that all new primary nodes built will be XGS-PON-enabled by October 2023. We should point out that existing customers on the network, which have older GPON ONTs, will continue to use those for now (an engineer normally needs to enter your home to upgrade these). CityFibre’s approach to all this involves a co-existence between the parallel running of their GPON and XGSPON based systems, as well as an inter vendor GPON/XGSPON ‘combo’ port solution.
Put another way, unless you order a package on their network that requires XGS-PON or are a new customer in an XGS-PON area, then you’ll probably continue to use the existing GPON kit until it reaches end-of-life. But ultimately this isn’t about launching new products today, but more about preparing their network for those to arrive in the near future.
Mark is a professional technology writer, IT consultant and computer engineer from Dorset (England), he also founded ISPreview in 1999 and enjoys analysing the latest telecoms and broadband developments. Find me on X (Twitter), Mastodon, Facebook, BlueSky, Threads.net and Linkedin.
« Councils Warn Digital Divide Holding Some Back in Broadband Slow Lane
Latest UK ISP News
- FTTP (6035)
- BT (3643)
- Politics (2722)
- Business (2442)
- Openreach (2407)
- Building Digital UK (2331)
- Mobile Broadband (2150)
- FTTC (2084)
- Statistics (1906)
- 4G (1822)
- Virgin Media (1769)
- Ofcom Regulation (1585)
- Fibre Optic (1470)
- Wireless Internet (1463)
- 5G (1411)
Nice to see this happen. I know that in Edinburgh, just a few weeks ago, they were still deploying GPON. Even with this announcement, they might still be running down the GPON stock?
That said, things don’t get exciting until new packages become available to order.
‘Even with this announcement, they might still be running down the GPON stock?’
Areas partially deployed will get GPON to at least some extent. They are indeed running down the stock of GPON ports and ONUs.
Future proofing to 10Gbps is all very well but they need to concentrate on delivering FTTP in the first place at speeds up to 1Gbps than once the majority of the UK can get FTTP than look at faster speeds.
10Gbps isn’t and shouldn’t be a priority at the moment, rolling out FTTP in the first place is.
The majority of the UK can get FTTP. https://www.thinkbroadband.com/news/9555-51-of-uk-premises-passed-by-an-fttp-network
They’re good to go.
Even if it couldn’t not their problem. Building out fibre is really expensive, upgrading it once it’s built isn’t.
I suspect that the upgrade to 10Gbps and their rollout plans are largely decoupled. I’d be surprised if their rollout plans were unaffected, with the exception of transitioning from GPON to XGS-PON.
Don’t assume that effort in one area of the business impacts another 🙂
*affected
They need to get the customers for what they have already laid, not just Cityfibre, but others as well. A lot of people are happy with that they have and see no reason to change.
51% of the country is supposed to have fibre available, so we are told, i think that is Openreach only, so I presume it is more than 51%.
Now I have sorted out the reason why my broadband was going so slow, need to get on and look at what plusnet can offer me.
thanks for the clarification of the technical terms not a lot of people would know 🙂
exciting stuff!
for those who want more understanding behind this tech, see links like these:
https://dgtlinfra.com/xgs-pon-fiber-based-technology/
https://us.hitrontech.com/learn/what-is-an-ont/
As nice as it is. Why would we need these speeds residental speeds of up to a gigabit are more than enough for most families!
It’s never about what you need today, it’s about what you might need, or want, tomorrow. Plus, everybody is different. Not so long ago somebody, somewhere, was saying the same thing about 1Mbps, then 10Mbps, then 100Mbps etc. Furthermore, being able to download entire games and computer backups in minutes or seconds, rather than hours, means less time waiting with a computer switched-on.
“GPON shares capacity of 2.48Gbps downstream and 1.24Gbps upstream between 8 to 32 users”
Chances are that currently most ONTs are capable of meeting the needs of the various speeds of customers, but three years from today with 32 users expecting 1Gbps will experience significant contention choking/slowdown. Upgrading to XGS will relieve that pressure, even if ISPs don’t offer 10Gbps connections through it.
However, it has also been reported that Cityfibre intend to increase their split ratio to 1:128 with XGS-PON.
In that case, the total download bandwidth is more or less the same, although it should result in less contention in the upload direction.
The main benefit is to Cityfibre themselves, who can deploy 1/4 of the OLT ports in new networks, and hence have smaller POPs with lower power consumption.
That and statistical contention works way better. Remember datastream ‘skinny pipe syndrome’?
the speeds we have today will easily do us for at least another 10 years !
Money is all they care!
@ne555
> Cityfibre intend to increase their split ratio to 1:128 with XGS-PON.
Nope. I think most XGS-PON networks run at 1:64, not many at 1:128
> and hence have smaller POPs with lower power consumption.
Wrong again on POP sizes, most are already built now!
@Mark Jackson, you say it is what we might need, but I agree with Sam, the amount of households that will need more than 1Gbit will be tiny, if any, unless they have a mansion with a load of people using the system to download large game files at the same time. Zzoomm here offers 2Gbit, I would like to know how many people have taken up that at £64 a month. a good price for that to be honest, and if there are a few wage-earning people in the house then between them, it is not a lot.
I know this is not a large sample of people and only from the few people I know that have taken FTTP, but most of them have gone for the 400-500Mbs speed, one have not even upped their speed at all, just gone for a 74Mb/s, they got fed up with their ISP getting on to them about going for fibre, so they went with what they had before speed wise.
It is ok paying for these high speeds if you get these high speeds, but since FTTP is shared, at some point in time when enough people are using it you may not get that speed.
Ad47uk, your first para seems to be saying “we won’t need higher bandwidth”, and your third paragraph seems to be saying “we’ll see a slowdown if we don’t have higher bandwidth” ?
Data use per household has been rising steadily for years. The advent of technologies like 4K TV and AI-asssisted services, smart homes, remote monitoring, remote control of car charging and electric heating etc etc will all keep driving that for the forseeable future, even if data use does have some finite limit. By the time the network has lost capacity, it’s too late, and when it hits capacity the important impact is not on the “speed” bandwidth demand so much as the latency as requests get queued (some Virgin Media users will be very familiar with this scenario). And what that means is that latency sensitive uses struggle. Gamers you may not care about, although if they’re paying their bills they might, but it would be massively disruptive to video call and conferencing, and to digital voice uses – so VOIP goes down the toilet, working from home becomes more difficult. That’s not so good, there’s even more troubling issues for businesses if capacity on their networks is constrained, but that’s perhaps beyond the scope of this discussion.
Networks have to keep ahead of the rising tide of data use by increasing bandwidth, because even if nobody “needs” or perhaps wants >1Gbps, rising aggregate data use is a reality, and failure to start building for that now will have extremely unwelcome consequences for customers that have nothing to do with download times, and would affect all customers on a constrained segment, irrespective of their contracted speed.
@An Engineer, CityFibre explicitly tout energy savings as a motivation for the XGS-PON upgrade. I parse that as a higher split, at least in typical residential cases.
@Iain: the power savings come about because XGS-PON can operate with lower power consumption than GPON. XGS-PON kit is also usually cheaper than GPON kit and a single fibre supports up to four times more XGS-PON ONTs compared with GPON ONTs due to higher bandwidth. All of this overcomes the fact that the cost of an XGS-PON ONT is typically higher than the cost of a GPON ONT.
Roger, quoting you:
‘XGS-PON kit is also usually cheaper than GPON kit’
‘All of this overcomes the fact that the cost of an XGS-PON ONT is typically higher than the cost of a GPON ONT.’
So which one?
GPON is cheaper per OLT port than XGSPON and there’s no ‘typically higher’ about an XGSPON ONT, it uses a burst laser that’s way more expensive than GPON optics, but I’m interested in how you reconcile those two statements.
‘@Iain: the power savings come about because XGS-PON can operate with lower power consumption than GPON.’
The power saving comes from having an XGSPON port serving more premises than a GPON port thanks to the higher capacity per port alongside better optics, so in the case of CityFibre half as many ports to serve the same number of premises.
‘It is ok paying for these high speeds if you get these high speeds, but since FTTP is shared, at some point in time when enough people are using it you may not get that speed.’
Your FTTC cabinet may have nearly 300 customers sharing a gigabit port back to the exchange. All broadband and most leased lines are potentially shared with software smoothing the experience out.
@Andrew G, I can see where you are getting the idea where I say we will get a slowdown without higher bandwidth. FTTC is a shared system, split between a certain amount of households, if by chance a lot of people that shares that band with does have 1Gb/s or more, and they are blasting their bandwidth, then other people may get less than what they have paid for.
I am not saying that no one needs higher bandwidth, but i am pretty sure, in fact I know, that people getting pulled into the higher speed with adverts and sale tactics.
i have seen it over years, people getting sold stuff they don’t really need. Years agao I used to see shops try to sell high end hi-fis to people who just wanted something to listen to their records and cassettes and really did not have to pay hundreds. TV’s were the same, in fact we still have that problem if you go into a shop. Buy this amazing £1,500 TV to watch your Coronation street on, instead of the £300 one that will do the same thing.
computers were the same thing, I have gone with a few people over the years for them to buy computers and even after telling the sale people what the computer is going to be used for they go directly to a machine with high-end graphics and a super-duper high-end fast processor. Great if you want to play games or some other graphics or CPU intensive task. but if all they want is a machine to browse the net, maybe do a bit of word processing, then a lower price machine is fine.
I used to prefer to build machines for people than for them to buy them already made, but some people just prefer to buy pre-built.
I understand that we need to build for the future and I understand that FTTP is that future, I have no problem with the technology, it is just another way of getting broadband into the house, I have used wireless broadband. If FTTP was in the house already, I would be using it.
My problem is the way it is being sold, the way people are being told yes, faster you need higher speeds.
I know that people who are on lower speed FTTP have already been blasted with emails to try and get them onto faster speeds, even if they still have months to go on their contracts, upselling is going to be the thing. I have said to these people to change their marketing preferences, that should stop it.
you’re saying that AI-asssisted services, smart homes, remote monitoring, remote control of car charging and electric heating need high speed data? No they don’t, most of these, they use a tiny amount of bandwidth most of the time. I have a lot of smart home devices and all they are doing for the most part is listening for a signal. My blink cameras use about 2Kb/s when in use and my door bell camera is around 3Kb/s when activated. Smart thermostats send a tiny bit of data. Car charging again take a tiny amount as well, it only need to send a tickle of data.
Not sure what you mean by A.I assisted services, we have that now to a certain degree with Google, Amazon and even Apple.
Video streaming uses a fair bit of bandwidth, but I did a test here last year and I had 3 devices pulling 4K video at the same time and playing some music on the Echo dot and they all worked fine, sure there was nothing left over to do much else, trying to download something at the same time was a no-go. but even with 36Mb/s second you still do multiple things.
Now I am not saying that people stay with that if they can get and want faster, if people are pushing their connection like that then yes get faster.
But if these providers want to sell FTTP then they have to come up with a reason why we need it. Many people I chat to, either online, family, friends, even colleagues when we chat about broadband see no reason to change from FTTC as the speed they have is fine for them.
No doubt providers will put up the price of FTTC, just like Plusnet have to try to push people onto FTTP.
As I have said before I am out of contract at the end of June, so I have to decide what I am going to do, if Plusnet can’t give me a decent offer on my FTTC service and want to push me to FTTP, then they will lose a customer. If I go for FTTP then I may as well go to Zzoomm, cheaper and a better network. But there are some providers giving good prices for FTTC, now broadband for one and a 12month contract.
Re the split ratio, remember this isn’t difficult to change with a little planning.
Eg take take 4 fibres(32 customers each) from each splitter to the OLT, combine these with a 4 port splitter next to OLT. You can then easily configure it as 32,64,96 or 128 customers per port.
Start at 128, and as usage grows and ports get cheaper reduce the split ratio down
An Engineer..
“Wrong again on POP sizes, most are already built now!”
It’s PON , not POP. What does them already being built have to do with anything?
The fact the PONs are built already doesn’t prevent them changing the split numbers.
They can easily increase the split further upstream.
@John
Read the thread! I quoted and responded to NE555. His comment was specifically on reducing POP sizes, not PON.
Personally I’m currently happy with my 35mbs download, 6mb upload, and although XGS-PON may be desirable for the future I don’t see the need for it now. I can see a case for 100mbs download with some of the streaming content now available, but that is easily doable with GPON dependant on the number of properties served by the OLT.
Technology advances incredibly fast and there’s nothing to say that manufacturers aren’t looking at developing OLT’s offering multiple PON services from a single OLT, who’s to say that a manufacturer like NOKIA wont produce an OLT with GPON, XGS-PON and 25G PON at the output simultaneously? Offering an easy upgrade path for any future bandwidth requirements. Another feature of DWDM is that you can offer bespoke wavelengths down the same fibre to accommodate business users if required, simply by connecting the service to splitters at either end. When I’m finally offered the move to fibre, I’ll probably go for the cheapest offering around 100mbs or faster, with the emphasis on price not speed.
That a bit like say :-
I am happy walking from Edinburgh to London and how long it takes.
I dont get the point of the building roads for cars, laying track for Trains, building Airports for Airplanes. They should all walk like I do.
You don’t need higher speeds, which is fair enough. But that’s not the only motivation. CityFibre say XGS-PON “will allow CityFibre to support more customers on each OLT port, enabling substantial network cost savings, reducing power use across its networks”.
“But that’s not the only motivation. CityFibre say XGS-PON “will allow CityFibre to support more customers on each OLT port, enabling substantial network cost savings, reducing power use across its networks””
I agree, the main motivation behind XGS-PON is probably to serve more premises from a single OLT. In the end none of it matters, most people will only pay for a service that serves their needs at the lowest price. As far as the argument about time to download games, that may become irrelevant if cloud gaming really takes off, where latency will be the issue rather than bandwidth; The closer cloud gamers are to the server, the better, with providers offering the most direct connectivity the likely winners.
‘I agree, the main motivation behind XGS-PON is probably to serve more premises from a single OLT.’
Absolutely.
@Ex Telecom Engineer
You said
“In the end none of it matters, most people will only pay for a service that serves their needs at the lowest price”
If they have any sense., but some people get caught up with the sales gunk that comes though the door or what their ISP sends to them via email or phone callks.
A lot of people go for a faster speed because they think it is needed. My other half did that with gigaclear, going for 1Gbit and now realising she really doesn’t need that speed. She is not stupid either, she is a vet with all the degrees that requires and more.
But you are right a lot of people will pay as little as they can for what they need. Myself included, like yourself I don’t need super duper speed, 36Mb.s is fine.
But i have been having a look around broadband providers and prices this morning in-between grass cutting, bike oiling and other stuff that should have been done weeks ago 🙂
If i stay on FTTC, I will end up paying more than what I am paying now on must providers, even now broadband have increased price of their 36Mb/s FTTC £28.50 after 4th of July, £25 for a couple of months before. They are all around the £25 or more mark now, I think they are trying to get people to jump to FTTP, now is slightly different as they don’t do FTTC. Onestream is around £22 a month for a 12 month contract, but I have no idea what they are like or even who they are.
I am thinking maybe FTTP is the way to go, because providers are making it more expensive for FTTC. it is just the hassle of the installation.
I am going to try something later, see if a provider will give the same deals on FTTC if I enter an address that don’t have FTTP from openreach.
If I do go for FTTP it will be with Zzoomm, but I need to ask them a few things first.,
Got to get on with some gardening and stuff.
Ex Telecom Engineer – exactly the reason I am NOT with BT. 35mbps and 6mbps would not cut it here at all. Glad it works for you, but pointing out that many it wouldn’t. You come across as old school BT thinking (FTTC that’s fast enough for the country my dear!)
“enabled 18 Fibre Exchanges (FEXs) across 13 of their UK towns and cities (for context – they’re currently live or building FTTP across 86 active locations).”
I’d like to know how many FEX’s there are across these active locations to put this into context, for example if they have enabled 18 in 13 cities and each city only has 1 or 2, that’s great, but if each city has 30, 18 over 13 is not so good.
As far as I know one fex per city. I can’t think of one with >1 off the top of my head
Lots of cities with way more premises than a single FEX can handle according to CityFibre – https://cityfibre.com/news/cityfibre-deploys-100th-fibre-exchange-to-power-its-full-fibre-rollout
‘A CityFibre FEX unit is roughly the size of a shipping container and able to serve symmetrical gigabit speed connections to up to 60,000 premises at full capacity.’
60/64,000 premises, probably 64,000 being the number given that’s 2,000 PON ports on a 1:32 split ratio, isn’t going to cut it for a city the size of Leeds, Bradford or Sheffield. Leeds has at least 4 FEX live right now.
Sick of reading all the press about how amazing city fibre are. I live in Horsham where city fibre dug up all our roads in March 2022. It’s now May 2023 & they still cannot provide us with a connection. So we are stuck on 36 Meg while City fibre tell the country they are installing 10gbps?
Cityfibre are amazing where I live stuck on overpriced unreliable unhelpful virgin for years now I’m free
Cityfibre have been incredible here, no sign of Openreach or VM doing anything to sort their shit out.
The article says that this is upgrades to the backbone, currently 8-32 houses will share 2.5 Gbps approx downstream.
Not good during the evening if 32 houses with students or kids try and download the internet.
I see the same on VM gigabit after 4pm and during the school holidays don’t try ang get a game off PSN or Steam as frankly you are only getting a share of the bandwidth.
So at least City Fibre is getting on with backend upgrades quite quickly.
It’s perfectly normal for CityFibre to take years to complete a build. They still have tons of streetworks going on in Horsham, they don’t rock up one month, dig each and every street, and offer service to everyone.
They dig a feeder network at first then branch out from that, so perfectly possible for there to be a dig for the feeder network then a dig to get from there to a street then an install of a cabinet and connection to the Openreach network for PIA, or dig down the street with toby boxes installed outside properties.
Openreach also have works all over Horsham as do Virgin Media. You are unlikely to be stuck on FTTC for long but building networks takes time as there’s only so much labour available in an area: someone in a town/city has to be first, someone last, and everyone else somewhere in the middle.
Yes someone is always first, middle and last, but it doesnt make much financial sense to take years to provide connectivity. Surely it is more efficient to cable a street and then connect it up, so that they can earn revenue from their infrastructure? Simply building thousands of miles of cable with no profit stream makes no business sense. You are outlaying all the costs up front and not getting any return.