It’s official: Ciena and Cisco named Verizon’s next-generation metro suppliers
Ron Kline, Principal Analyst, Intelligent Networks
Some of the biggest buzz at OFC 2015 was around the long-awaited announcement by Verizon that it has selected Ciena and Cisco to provide packet optical equipment for its next-generation 100G metro network in the US. This will be the first large-scale deployment of flexible-grid colorless, directionless, and contentionless (CDC) ROADMs.
The news is not unexpected, as it was rumored in January, but the official word is good for both Ciena and Cisco. While the rumor mill has pointed to Cisco as the bigger winner in this deal, in fact it’s too early to tell. Our contacts at Verizon indicated to us that the two vendors’ relative share of the project will be guided by ongoing testing, support, and performance. The Verizon deployment will not impact vendor revenues until 2016.
The announcement marks a significant milestone for packet optical systems, which Verizon has been pushing the industry towards for several years. The platforms selected by Verizon (Ciena 6500 and Cisco NCS 2000/4000) incorporate the latest converged packet optical capabilities, including coherent 100G, OTN/packet switching at the electrical layer, and CDC ROADMs at the photonic layer. In addition, Verizon plans to use circuit emulation technology to support the migration of the thousands of TDM ports in its network to Ethernet.
Ciena has been a long-haul optical supplier to Verizon for many years. The win will expand the company’s addressable market and also increase cross-sell opportunities for its 5400 OTN switch. Cisco is not currently an optical supplier to Verizon except for some enterprise customer applications. The win culminates a multiyear development cycle largely driven by the desire to break into the Verizon network. Kudos to the winners.
Component vendors to benefit from Verizon’s next-generation 100G metro deployment
Daryl Inniss, Practice Leader, Components
Verizon this week announced the selection of Ciena and Cisco as suppliers for its advanced 100G metro network.
The deployment is rich with new technology and provides an excellent commercial opportunity for component suppliers. Component products in three areas will be needed: coherent optical transmission at 100G; flexible colorless, directionless, and contentionless (CDC) ROADMs; and packet switching.
At the beginning of the deployment, discrete components will be used for coherent transmission at 100G and beyond. These include modulators, lasers, and integrated receivers. JDSU is in a strong position to supply the modulators, and NeoPhotonics will benefit from the laser demand. We also believe NeoPhotonics will win some of the receiver business, along with other suppliers. Transmission modules, most likely the analog coherent optical CFP2 transceiver, will be introduced in the network during the deployment cycle.
The flexible CDC ROADM has been a challenging product segment over the past few years. Revenues have been flat to contracting even as investment was needed to deliver higher-density solutions and increased functionality. Finisar, JDSU, and Nistica are well positioned to benefit from Verizon’s deployment.
Packet switching is also needed, as Verizon plans to support its legacy services on an IP network. Circuit emulation technology, for example, allows transmission of time division multiplex service over a packet infrastructure. Expect Xilinx, Altera, PMC Sierra, and InPhi to benefit from the demand for FPGA and OTN chips.
Verizon’s deployment is a multistage process that will take years to fulfill. Component vendors must continue to innovate and deliver leading-edge technology to succeed commercially.
M/A-Com adds support for 10G PON BOSA-based design, following the movement to BoBs (BOSAs on board)
Julie Kunstler, Principal Analyst, Components
M/A-Com issued a slew of press releases during OFC, including a product announcement for its new M02180, an optical interface chip for 10G PON ONTs/ONUs. The M02180’s laser driver is integrated with a limiting amplifier and clock and data recovery (CDR). The M02180 laser driver can be configured to support 10G FTTx BoB applications.
The vast majority of 2.5G GPON ONT/ONU designs today are based on BoB designs. As Ovum discussed back in 2011 in BOSA-based PON Customer Devices Gaining Share (TE003-000495), the cost of a BOSA (bidirectional optical subassembly) was 30–40% lower than that of an optical transceiver module. This cost advantage is what helped BOSA-based PON get off the ground, and technical/commercialization challenges have also been resolved, so they’re now mainstream.
The adoption of 10G PON equipment is taking much longer than many expected, including Ovum. One delay is the cost of 10G optics, reflecting the classic “chicken-egg” theory. Demand will bring down the costs, but the costs are currently too high to attract volume.
From meetings at OFC, we believe that 10G PON shipments are increasing as several service providers (telco and cable operators) move ahead with 10G PON FTTx deployments. While 10G ONT/ONU shipment levels are forecast to remain under 500,000 in 2015, the ability to use BoB designs for 10G helps to lower the cost regardless of volume. This should help solve the chicken-egg problem.
Cisco allocates booth space to FTTx PON
Julie Kunstler, Principal Analyst, Components
Cisco’s demonstration of FTTx PON equipment is noteworthy, but not because its PON equipment is brand new. We believe that Cisco has been quietly supporting Google Fiber’s FTTx deployment along with several other vendors. Now we see Cisco’s open commitment to the PON market.
Why was Cisco reluctant? PON equipment is for access, and access equates to lower margins versus enterprise equipment, according to conventional wisdom and lots of anecdotal evidence. Second, Cisco is the leader in CMTS equipment, and marketing PON equipment might cannibalize those sales and lower margins. Third, Cisco is a strong proponent and player in point-to-point Ethernet solutions, while PON is point-to-multipoint.
Why now? Ovum believes that Cisco is seeing several market trends that acted as triggers to increase internal development and sales efforts:
Several large US cable operators have begun to deploy PON for residential and business services in high-density areas. Ovum believes that Comcast and Time Warner will deploy 10G symmetrical PON equipment this year. Cisco does not want to lose equipment revenues to the major MSOs. Cable operators don’t just want PON, they need it.
Next-gen PON was designed to support the enterprise sector, Cisco’s core market. While existing PON equipment is supporting SMBs (small-to-medium businesses) and mobile backhaul, next-gen PON has technical advantages, including higher bandwidth and the ability to assign lambdas to specific applications and/or customers.
There are skeptics who say that Cisco is too late to the market. Cisco is too late to win in China, but that was not its goal. Cisco is not too late to support next-gen PON deployments.