fibeReality fully expects Ericsson/Calix to be the primary beneficiary of NG-PON2 becoming the de facto standard for much of the world. We anticipate that the vendor combo will take the lead on both 5G wireless and supporting optical backhaul/fronthaul equipment for Verizon, which is in the process of quietly setting up next-gen infrastructure well beyond its publicized trials. The service provider is fully committed to the realization of NG-PON2, which would represent the first time in the history of optical communications that such a large disparity at the outset in first cost would not be a prohibitive factor in moving forward on a solution. Despite any rhetoric to the contrary, Verizon has its whole future riding on the success of 5G, and other access alternatives are simply inadequate to completely provide all of the desired functionality.
While overcoming the high cost of NG-PON2 will perhaps be Verizon’s biggest challenge ever, it has demonstrated in the past that it is tremendously unlike other incumbent ISPs anywhere in the world, including in dealing with exorbitant expenses effectively. While being attacked unmercifully last decade for its FiOS deployment, it managed to pass as many Fortune 2000 customers as possible in order to subsidize the CAPEX for residential services. (It has only been recently that Verizon partially admitted to its prior strategy, which was developed back in the 1980s at the original Bell Atlantic, if it received the required regulatory relief.)
Although it is hardly an optimal time to be a public network transport system vendor, both Ericsson and Calix are particularly suffering with subpar performances. Verizon will be sure to take advantage of their plight to obtain massive price concessions. We think it is very conceivable that Calix, which is already taking a hit in being forced to be part of an OEM arrangement, will lose money with sales to Verizon in exchange for becoming the perceived leader of NG-PON2, and make up the margin loss with substantial sales of the gear to other operators. At some point, volumes should reach a point in which the first cost will no longer be an issue anyway.
Given the long-time prejudice against Nokia by Verizon, we do not expect it to play much of a role in providing gear to the ISP in its 5G network, other than perhaps there is a situation in which demand exceeds supply for ONTs. Actually, Verizon’s willingness to sacrifice its staunch resistance in general to disaggregation in optical networks with its OpenOMCI concept shows the seriousness of purpose in which it is committed to making NG-PON2 cost-effective.
If absolutely necessary, Verizon will probably adopt other approaches from the hyperscale data center operators in order to get the cost of NG-PON2 down, whether it is making investments in optical component suppliers or as a last resort, even doing in-house production. However, we do not foresee the inevitability of such extreme measures. For example, we are counting on the cost of tunable lasers coming down as more componentry vendors look to take advantage of this massive opportunity over the next several years.
On a final note, despite the absence of much discussion at ECOC 2017 about NG-PON2, Vodafone is supporting the standard. Of course, Vodafone used to be a business partner of Verizon, and knows firsthand that when the US-based company sets its mind to something, it is likely to get accomplished.
[written by Mark Lutkowitz]