Article Info

The Future of FTTH in China – Part III

This is part III in a continuing series. Part II can be found here.

We believe China Telecom (CHA), as well as other Chinese actors, will choose GE-PON. They are motivated by factors that transcend technical specifications, factors that matter most to the selection process.

The Chinese Tao

Contrary to popular belief, both government and business in China do not function through rigid top-down execution of initiatives. China is a confederation of provinces and townships with varying amount of influence, and each is slowly coerced into following a high level dictum through social engineering mechanisms foreign to most western businesses.

There is an excellent chapter in the book One Billion Customers (Wall St. Journal Press) that provides the story behind the PHS ‘Handyphone’ system and how it gained popularity contrary to the desires and dictations of Beijing. With PHS, there were striking technical and economic advantages that overwhelmed the states desire for central planning.

Aside from the reams of marketing material available, the G-PON protocol has no dramatic technical or economic advantages over GE-PON. There are indeed some rogue states within China that have deployed G-PON in limited amounts, these are well highlighted in Flexlight press releases.

This is why GE-PON will win in China.

  1. Component Availability – There are at least seven silicon vendors who have announced GE-PON chipsets but only four for G-PON chipsets. More importantly, there is only one G-PON chipset that is perceived as market available.

    This is incredibly important as Chinese vendors are well known for not entering a particular market until a robust supply chain exists. Multiple sources for components allow them to use their large volumes to play vendors against each other and drive radical cost reduction. Huawei is particularly good at this, and since no secrets exist in China, the Huawei price becomes the China price.

    Optical component vendors have already endured this pricing bloodbath, and the relatively few optical specification differences between the two technologies should not affect China’s standard selection. ADVANTAGE: GE-PON

  2. Cost – Contrary to the marketing materials pushed by G-PON vendors, a GE-PON system is cheaper than a G-PON system. You will see charts showing that at a certain bandwidth per user and certain number of subscribers per link G-PON gets cheaper, but in reality none of these matters – the startup costs for GE-PON are cheaper today. This is due to the fact that there will be several million GE-PON nodes deployed this year, and virtually zero G-PON nodes deployed.

    The relative low cost of labor in China, and the resulting high percentage of costs assigned to equipment makes hardware pricing a greater issue than in the west. The lower startup costs of GE-PON make it a clear winner in China as a result.

    Also, in the components business, volume is as important a cost lever as technology. Fixed costs (R&D, Qualification, Fab Capex) tend to dominate therefore the incremental cost of another component decreases as volumes increase. GE-PON will have much higher volumes for the forseeable future, and therefore lower structural costs. ADVANTAGE: GE-PON

  3. Maturity – NTT has deployed millions of GE-PON nodes and trailblazed through innumerable technical challenges. Verizon is only now beginning a similar effort with B-PON. G-PON has yet to encounter any problems because it hasn’t been deployed yet at all.

    This is compounded by the lack of proven G-PON interoperability among vendors. Telcordia is holding its very first G-PON interoperability event in 2006, while GE-PON has underwent informal interoperability testing in NTT labs for several years. The presence of a single large customer, NTT, introduced an informal interoperability requirement that forced the marketplace to quickly converge GE-PON requirements. ADVANTAGE: GE-PON

  4. Flexibility – The IEEE 802.3ah spec is relatively lightweight, with many of the features written as optional add-ons. GE-PON equipment initializes communication by negotiating with the known standard, then begins to query connected devices on their capability to support addendum requirements. GE-PON chipset vendors are accustomed to making modifications and in many cases have designed their architectures around such flexibility. Implementations can be custom but still meet the specification. The ITU spec does not allow for such flexibility, and it turns out that flexibility is something China Telecom is looking for. ADVANTAGE: GE-PON

These four factors could apply to the decision of any telecom actor, not just a Chinese one. At this time it is unlikely that the large American and European telcos will use anything but the ITU B-PON and G-PON standard. It also appears virtually certain that China will chose GE-PON based on the structural economic advantages it brings to the table.

However, the most compelling evidence that China will select GE-PON is based on the actions of it’s largest carrier. There is a quiet and coordinated effort by China Telecom to enhance the IEEE GE-PON specification to meet the needs of their network, something informally referred to as C-PON. This effort and the companies involved will be examined in Part IV of the series.

Continue reading part IV (Link inactive until June 20th)

Companies Mentioned
China Telecom Corp Ltd (NYSE CHA [ADR])
Flexlight Networks – Private
Nippon Telegraph and Telephone Corp (NYSE NTT [ADR])


Comments are disallowed for this post.

  1. I think that the following needs a little clarification:
    “… China is a confederation of provinces and townships with varying amount of influence, and each is slowly coerced into following a high level dictum…”

    My concern is this “high level dictum” bit. The fact is that most orders issued by the Centre never get out of the Compound!

    Instead, local officals increasingly are local emporers. Witness the recent spates regarding land/property, or indeed pollution control.

    There may be “high level dictums”, but they stay in the clouds (except as regards police and military matters….). The simple point is that China’s leaders from the provinces down are growth-obsessed, all the moreso the next 18 months when one of China’s largest overhauls of its civil service is about to occur…

    Posted by Enzio von Pfeil | June 13, 2006, 5:59 PM
  2. Trying to distill a description of how Chinese government works into a single paragraph is a challenge.

    Posted by Andrew Schmitt | June 13, 2006, 9:14 PM
  3. I agree with the current market status regarding GPON and GE-EPON in China, but telecom operators like China Telecom(CT) and China Netcom (CNC) still think GPON will be the final winner because they believe: 1)GPON is ITU standard with many advantages over GE-PON2) GPON price will be the same as EPON price in near future after GPON ASIC is available and business volume groups up. 3)China Telecom and CNC will not deploy GE-PON in large scale in near term because IPTV can’t be deployed in large scale because of many specific difficulties in China, this will give CT and CNC enough time to wait for GPON solution maturity.
    In addition, mobile carriers like China Mobile are think of using GPON in mobile BTS backhaul application, this will also be a big market and the price is not sensitive.
    So, GPON still look good in China.

    Posted by Simon Wang | June 14, 2006, 11:08 PM
  4. It is IEEE 802.3ah’s flexibility that does not lend itself to interoperability. As you mentioned “while GE-PON has underwent informal interoperability testing at NTT labs for several years” there’s still only one silicon vendor at NTT. On the other hand, ITU-T G.984 has undergone a very successful interop at Telcordia with 10 participants. There’ll even more interops sponsored by FSAN and ITU in 2006 with even more participating. In such a short time, GPON has come a long way due to its well defined standard. This also opens the way for multiple silicon vendors. The success of DSL is a result of such interop activities lead by international standard committees and not carriers.

    Posted by Dan Parsons | June 21, 2006, 4:58 PM
  5. I think this all comes down to the technology that is ready for prime time to meet the 2008 Olympics. GPON implementations will just ready to go into the RBOC labs in September, and you can add a good year before they are ready for inital deployments. I would think the Chinese would need a technolgy platform that is ready for primetime by Summer of 2007.GEPON is the clear winner in this regard.

    Posted by drw | July 3, 2006, 2:48 AM
  6. I wish I had a dollar for every time someone mentioned the Beijing olympics and massive telecom capex.

    How about some hard data showing a link?

    Posted by Andrew Schmitt | July 4, 2006, 11:56 AM
  7. ‘Beijing olympics and massive telecom capex’ is just a hype of those hungry equipment vendors. In fact, there will be mere small telecom capex for wiring those new stadiums in Beijing.

    Posted by David King | July 5, 2006, 12:17 AM
  8. Trackbacks / Pingbacks

  9. FTTHblog » Parts 3 and 4 of Nyquist Condition China FTTH Series | June 21, 2006, 12:27 PM
  10. The Future of FTTH in China - Part V at Nyquist Capital | November 10, 2006, 12:20 PM
  11. The Future of FTTH in China - Part IV at Nyquist Capital | November 10, 2006, 12:36 PM