Article Info

The Future of FTTH in China – Part V

What we’ve learned

  1. Part One – A radical change in demographics is creating an optimal situation for a fiber build out in China
  2. Part Two – GE-PON has taken hold in other countries with similar demographics
  3. Part Three – The economic and market conditions appear to favor GE-PON
  4. Part Four -The activities by the largest Chinese Telecom actor reinforces this choice

In this fifth and final installment we will examine what questions remain that prevent accurate forecasting of PON deployments in China.

We feel the greatest question is whether the apartments, the overwhelming bulk of new construction in China, will have fiber connections to each unit. This has a dramatic impact on the amount and type of equipment that is deployed, and therefore an essential element in determining who will benefit.


Fiber to the Home (FTTH) is a grossly abused term when used in reference to multi-tenant dwellings. Many would classify all the following scenarios as being FTTH, when many are clearly Fiber to the Node (FTTN).

  1. Each apartment is served by an individual fiber and has an ONU installed.
  2. Each apartment is served by a copper based VDSL connection terminated in a wiring closet on that floor of the building. A PON network may or may not be used to serve this closet node.
  3. Each apartment is served by a copper based VDSL connection terminated in the basement. A PON network could be used to serve the basement node, though it is unlikely if more than 4-8 subscribers are served.
  4. The entire apartment complex is wired for Ethernet using off-the-shelf hardware typically used in an Enterprise environment. The WAN uplink in the basement could be anything from state-of-the-art Metro Ethernet to bonded T-1’s using ML-PPP.

Scenario one requires an ONU chipset for each customer. Scenario two requires per-customer VDSL silicon and less PON silicon. Scenario three doesn’t even use fiber to serve individual customers, and those FTTH subscribers are really using DSL SLICs instead. It’s clear that not only are some FTTH customers really Fiber to the Node, but that there is not a 1:1 relationship in subscribers to components consumed. Existing market research reports do a horrible job of clarifying this.

Considering the vast majority of Chinese customers will be in apartments, determining the impact of multi-tenant dwellings on FTTH deployment is of crucial importance.

In Japan, roughly two in five of the NTT FTTH subscribers are apartment dwellers. NTT runs fiber to some, and uses VDSL to serve others. Landlords or Apartment Associations sometimes refuse to allow NTT to install the fiber, though such resistance is slackening as fiber connections in Japan have become the new ‘Cool Thing’ to have, and property owners don’t wish to be left behind their peers. While exact numbers are difficult to obtain, it does not appear VDSL has been used in substantial numbers.

However, China is a whole new ballgame with a different set of economic constraints. It is very unclear which of the above five scenarios will dominate.

Regardless of this uncertainty, we must have a model for forecasting FTTH in China. It is a straightforward model build on the fact that most new construction in China will have fiber installed, and that the rate of new construction will be proportional to long-term migration trends. Adjacent to this, we expect overall (including FTTH) penetration to follow historical adoption trends among the urban population of China.


  • PON deployments should track linearly with the urban migration trends now dominating Chinese demographics. New construction is used to expand the capacity of a city due to migration, with existing residents trading their worn low end residences for newer, nicer dwellings.
  • VDSL deployment will track linearly with the growth of broadband penetration in China, derated by a co-efficient related to urban migration.


This report is a high level summary of our thoughts on China’s impact on broadband deployment and the resulting impact on component and equipment suppliers. Please contact us directly for more in-depth analysis and actionable investment conclusions.


Comments are disallowed for this post.

  1. You seem to be of the opinion that GE-PON will dominate in China. How do you assess Teknovus’ chances there against PMCS at the major Chinese carriers? Their intense focus solely on E-PON, unlike PMCS, seems like it has an opportunity to pay off there. Also, do you think smaller outfits like Fiberhome have any chance to win significant FTTH business there against the likes of Huawei and ZTE? Thanks.

    Keep up the good work on your blogsite. It’s one of the best out there.

    Posted by Niles | February 1, 2007, 4:03 PM
  2. I don’t think it is possible to call the horse race between PMCS and Teknovus. Each has unique advantages – PMCS (Passave) has an installed base in Japan and deployment credibility. Teknovus is un-encumbered by customers and likely can react better to the shifting requirements in China. PMCS can also bring in Voip functionality that will be crucial for integration in the gateway.

    As for PMCS, until recently Passave was all GE-PON. G-Pon is a new thing for them. PMCS is very focused on GE-PON.

    This is what I do know: Chinese customers never single source a vendor. If they can’t have two wives they don’t get married. I think both vendors win.

    UT Starcom has a solid history of winning business within China. The company was built on their success in the PHS business, success that was won through the kind of negotiating tactics I never learned.

    It’s way too early to tell who will or who won’t win.

    Posted by Andrew Schmitt | February 2, 2007, 4:50 PM
  3. Trackbacks / Pingbacks

  4. FTTHblog » Issues of Fiber Reach in Asian Multi Dwelling Units | June 27, 2006, 12:34 PM
  5. » Could the Bloom be Off Wireless? | August 15, 2006, 11:39 AM
  6. State of the Photon - Global FTTH Activity at Nyquist Capital | May 31, 2007, 5:02 PM