Dosis 3.1 Upload increase?

[Anonymous]

Will virgin ever have a 1:1 down and up speed/connection with dosis 3.1 to match the competition? From what iv seen dosis 3.1 can easily handle it and upload is also just as important nowadays.

[legacy1]

---

@Anonymous wrote:

Will virgin ever have a 1:1 down and up speed/connection 

---

In 10 years time maybe.

[Roger_Gooner]

Any competitors who provide symmetrical speeds are tiny players which VM doesn't have to worry about.

[Timwilky]

VMs model is so 20th century. These days of cloud storage, home working. Their asynchronous model of all people want is download for porn and gaming no longer works

[Andrew-G]

DOCSIS 3.1 can't handle 1:1 (full duplex), unless VM chose to take the relevant standards out of DOCSIS 4.0 and implement them over the top of the existing D3.1 installation, and replace every customer's hubs with a model that doesn't yet exist.  Technologically and commercially that would be bonkers, so that won't be happening.  VM recently announced a plan to change the entire UK coax network to different technology that will see all coax replaced by 2028 with a full FTTP solution that doesn't use DOCSIS, and that would potentially allow full duplex.  VM's history is that they only ever launch new speeds and capabilities when they believe the market is ready for them, and willing to pay the full cost, and they only make major network changes after previous investments are amortised - if that's the case with the FTTP changes, then the first non-trial customers would not see any of these changes until 2024 at the earliest, and then the national roll out would be progressive across 2024-2028.  You might see a further bump in download/upload speeds in the meanwhile via D3.1, to perhaps 1.5 or 2 Mbps, and maybe 100 Mbps upload, but full duplex, nope. 

[gitty]

Technically, 3.1 can offer symmetrical gigabit, but it requires a very large and expensive upgrade. The high-split option in 3.1 can take the upstream capacity above 1 Gbps. How much depends.

It may also be a requirement to do a lot of node splits (to get the service group size down) for symmetrical gigabit to be a viable option.

It can be done, but VMO2 doesn't seem to want to go down that road.

Other operators out in the world are doing it (the high-split), with a future upgrade to 4.0 in mind.

[Roger_Gooner]

As VM has announced that the whole of its HFC network will be upgraded to fibre and using XGS-PON the capability will be there for symmetrical speeds. That's what the S means in XGS-PON. There is, of course, no need to waste money on symmetrical for everyone, just implement it for the small minority who need it.

[Andrew-G]

There is, of course, no need to waste money on symmetrical for everyone, just implement it for the small minority who need it.

I would be very surprised if the customers who need symmetrical speeds would be able to afford the costs if they were the only ones paying, since "implementing for the minority" across a network will cost near enough the same as for everybody.  And looking at the D3.1 upgrade that has cost in the region of what, £4bn?, we're all paying for that, despite the fact that only a tiny minority of customers are using the 1 Gbps speed that D3.1 enables.  

I suspect that D3.1 was a singularly unwise investment choice for VM.  It has put off the conversion to FTTP, symmetrical capacity, and the abandonment of DOCSIS by at least four or five years now.  Had VM made the commitment even sooner, then much the Lightning build out could have been done (albeit with complications) as PON rather than the FTTP RFoG camel.  In that scenario, 20% of VM's network would now already be operating as PON, the D3.1 costs and skills could have been used to get PON to perhaps another 20-30% of the network, and by starting ahead of Openreach's full scale buildout, VM could have had first dibs on the supply chain for PON.  Is that hindsight, or simplistic?  I think not.  I've worked on IT strategy, and it's always about making complex decisions for the long term, and the one thing that never works is trying to enhance "sunset" technology when it clearly doesn't have a long term future.  If you have a monopoly it might work, but with Openreach plotting near national FTTP, VM's monopoly of high speeds was clearly time limited. 

I would guess that like so many businesses, the draw for VM's management was of a perceived safe and easy choice that they felt they understood and where short term considerations trumped the longer term benefits of choosing the right technology.  VM's technologists may not have been quite as keen as management to do down the D3.1 cul-de-sac, but poor tech or customer service decisions are inevitable when corporate culture and the board are completely dominated by sales and marketing b****s******s.  All credit to the visionaries who've finally persuaded the board that D4.0 is not the way forward.

[Roger_Gooner · Accepted solution]

In 2015 VM trialled FTTP in Cambridgeshire and Leicestershire using RFoG, got it working satisfactorily and started to roll it out. In fact quite soon afterwards most new installations have since been FTTP with new HFC installations actually being infills rather than brand new.

The upgrade from DOCSIS 3.0 to 3.1 started in 2016 after Cablelabs began to issue certifications for modem vendors. The upgrade to DOCSIS 3.1 was absolutely the right thing to do as it makes sense to do this (good luck in getting 10Gbps download out of DOCSIS 3.0) and VM has a major advantage of using the same hubs and TV boxes for both the HFC and FTTP networks.

However it's now 2021, not 2015/16, plenty has changed as VM has merged with O2 to become VMO2 with some 46 million customers, the VM-Vodafone deal is unwanted and strictly time-limited for the mobile network and VMO2 now wants to greatly expand its fixed network and sign up some major ISPs such as Sky. This wholesaling of the network will clearly work best if it's all-FTTP and, having done the sums, VM reckons that the cost per premise of upgrading to FTTP (£100) compared with £60 for future upgrades to DOCSIS 4.0 for the 14.3 million HFC premises is affordable.

So, in several years from now we'll see the DOCSIS network over HFC and FTTP phased out and in its place with be XGS-PON over FTTP. This is entirely sensible for VMO2 but, if it weren't for the merger with O2, it would have been sensible to stick with the original strategy of getting more out of the network with DOCSIS 4.0 (although I do believe that RFoG would have been replaced with EPON for broadband).