RBOCs Aim for 10M GPON Subs

The nation's three largest phone companies -- AT&T Inc. (NYSE: T), Verizon Communications Inc. (NYSE: VZ), and BellSouth Corp. (NYSE: BLS) -- are anticipating that there could be as many as 10 million customers served by gigabit passive optical network (GPON) connections by the end of 2011, Light Reading has learned.

That forecast, disclosed in the three-carrier request for proposal (RFP) put out in November, asks prospective vendors to assume that GPON penetration could reach as many as 9 million residential consumers and 1 million businesses within the next five years.

One analyst reports that Tellabs Inc. (Nasdaq: TLAB; Frankfurt: BTLA), Alcatel (NYSE: ALA; Paris: CGEP:PA), Motorola Inc. (NYSE: MOT), and Hitachi Telecom (USA) Inc. are on the carriers' shortlist to supply the gear. "Alcatel, similarly, may have an advantage at Verizon as it is the incumbent access equipment (DSLAM) vendor inside the account," writes George Notter, analyst at Jefferies & Co. Inc. . "While there seems to be pros and cons to selecting Tellabs, we're expecting that Tellabs will win some business as either a primary or secondary source at Verizon (we're leaning slightly into the camp that says Tellabs will be a second source vendor)."

According to Light Reading's sources, if Notter's report is correct, it signals a huge disappointment for Ericsson AB (Nasdaq: ERIC), which bid with Entrisphere Inc. ; Nortel Networks Ltd. , which bid with Huawei Technologies Co. Ltd. ; Lucent Technologies Inc. (NYSE: LU), Fujitsu Ltd. (Tokyo: 6702; London: FUJ; OTC: FJTSY), and Calix Inc. (NYSE: CALX). Though one source close to the situation tells Light Reading that Huawei may be kept in to help drive pricing down. (See Fujitsu Preps GPON Plans, Nortel, Huawei Bid on GPON, and GPON RFP Weighs In.)

But what of this 10 million customer forecast? Those aren't huge numbers considering how many customers those carriers serve now. But for GPON, a ramp up to 10 million connections in 5 years is substantial. And those RFP forecasts include every kind of deployment scenario -- new (greenfield) deployments and overlay deployments -- and every kind of end user connection (VDSL, Ethernet, or FTTP).

Table 1: GPON RFP Deployment Forecast
BUSINESS CUSTOMERS 40,000 200,000 9,000,000
RESIDENTIAL CUSTOMERS 360,000 1,800,000 1,000,000
TOTAL GPON CUSTOMERS 400,000 2,000,000 10,000,000
Source: RBOC forecasts, Light Reading sources

Table 2: GPON Connection Breakdown for 2006/2007
TYPES OF ONT (Optical Network Terminal) CONNECTIONS DEPLOYMENT VOLUME for 2006/2007
Single Family Unit 306,000
Multi-Dwelling Unit (VDSL- and Ethernet-connected ) 7,600
Single Business Unit 3,600
Multi-Tenant Unit 3,600
Source: RBOC forecasts, Light Reading sources

Table 3: GPON Connection Breakdown for 2009
Single Family Unit 1,530,000
Multi-Dwelling Unit (VDSL- and Ethernet-connected ) 38,000
Single Business Unit 18,000
Multi-Tenant Unit 18,000
Source: RBOC forecasts, Light Reading sources

A few years ago the major service providers put out a mid-range forecast when requesting proposals from vendors related to BPON gear. At that time, the three phone companies cumulatively were expecting to serve about 8.1 million BPON connections by the end of 2008.

They're getting pretty close. Verizon alone says it will pass 6 million homes with its FiOS (BPON) network by the end of this year. And, in Verizon's case, after it passes those homes, an upgrade from BPON to GPON is certainly possible, depending on consumer demand and equipment costs.

Speaking of costs, another interesting tidbit from Notter's note is that the higher GPON volumes may be a trigger to bring fiber access pricing to its lowest point ever. "Our contacts indicate that pricing on this GPON RFP will certainly be at or below current pricing for BPON systems," Notter writes. "We found these anecdotes a bit surprising as GPON components will certainly have higher cost points than BPON components."

The GPON connections called for in the most recent tri-carrier RFP will provide voice, video, and data connections with nearly 2.5 Gbit/s of downstream bandwidth and 1.25 Gbit/s of upstream bandwidth that can be split up to 32 times, depending on the network configuration.

According to sources familiar with the RFP, the carriers are asking vendors for FTTP systems that provide a "full suite of competitive video services". This includes the ability to provide "sustained delivery of up to 100 Mbps of Ethernet frame payload... for television streams at the ONT." And they want each network-facing gigabit Ethernet port on the OLT to be able to manage 4,096 simultaneous and unique multicast sessions, limited only by the total interface bandwidth.

Continue to page two

1 of 2
Next Page
<<   <   Page 2 / 2
MorningWd 12/5/2012 | 4:02:26 AM
re: RBOCs Aim for 10M GPON Subs Schmitt,

I completely agree that the largest barrier by far is the physical construction costs of deploying the fiber. I would argue that GPON will quickly replace BPON, not for "market need and applications," but for a movement away from ATM. Why continue to deploy an ATM-based FTTH technology if GPON creates an IP-friendly replacement version? Let me restate "replace" with cap BPON deployments.

Mark Sebastyn 12/5/2012 | 4:02:26 AM
re: RBOCs Aim for 10M GPON Subs Guys:

If there is anythiing you can bet on, it's the slow migration to new technologies by the bells. Realistically, why is GPON needed? B-PON and a secondary wavelength supply sufficient bandwidth for existing applications.

The big barrier to entry is fiber in the ground. It's relatively easy to then upgrade to G-PON. By the time G-PON bandwidth is needed, my guess is that functionality not anticipated today will be needed in the ONU.

Therefore, until the market need and applications exist that justify GPON, B-PON will continue to be deployed.

People are attached to GPON because it is the only hope hardware vendors and chip vendors have to unseat the incumbents and play the FTTH card within the investment community.

More commentary on incumbency:

RTL Rules 12/5/2012 | 4:02:15 AM
re: RBOCs Aim for 10M GPON Subs I would argue that the same thing will happen with GPON when carriers "inevitably" move to EPON.

'Course I don't have a no crystal ball, so this is just IMHO.


ps I'd like to have that crystal ball to see whether Tellabs' GPON delay will affect their chances in the RFP!
wap545 12/5/2012 | 4:02:06 AM
re: RBOCs Aim for 10M GPON Subs Do not count out Moto in this process. They are the biggest player in the home video market (end to end) with their DCT Set top Box and video head end systems and have a major position in the APON FTTH space today. Moto should also dominate the Home IP based STB space and are actively developing a GPON product based on their Quantum Bridge platform.

Mark Sebastyn 12/5/2012 | 4:02:03 AM
re: RBOCs Aim for 10M GPON Subs Motorola may do OK in the FTTH business, but it won't be because of the STB business. The STB is going to migrate to a commodity PC of hardware, just like the PC did, with an OS from someone like Tivo. In the case where the STB is not commoditized, it will be vertically integrated vendors who deliver a home entertainment platform like MSFT/AAPL/SNE.

Cisco will be stuck holding the bag on Scientific Atlanta.

rs50terra 12/5/2012 | 4:01:54 AM
re: RBOCs Aim for 10M GPON Subs There are two reasons why they moved to GPON.

The first is that BPON doesn't have enough bandwidth to support IPTV. It can support broadband video using RF overlay, but this is the same service I get today on my cable and the only reason one would switch to a RBOC is getting lower fees and/or better programs. With GPON the RBOCs have enough BW to deliver IPTV.

The second reason is that BPON is ATM based and the RBOCs have finally seen the light and try to move to an all Ethernet network.
rs50terra 12/5/2012 | 4:01:53 AM
re: RBOCs Aim for 10M GPON Subs I think there is at least another vendor on the short list.

Guess who?
trzwuip 12/5/2012 | 4:01:52 AM
re: RBOCs Aim for 10M GPON Subs Why will they need another 1? 4 is enough..
<<   <   Page 2 / 2
Sign In