100 Gigabit 'unstoppable'

A Q&A with Andrew Schmitt (@aschmitt), directing analyst for optical at Infonetics Research.

"40Gbps has even less value in the metro than in the core"

Andrew Schmitt, Infonetics Research

A study from market research firm, Infonetics Research, has found that operators have a strong preference for deploying 100 Gigabit-per-second (Gbps) technology as they upgrade their networks.

Infonetics interviewed 21 incumbent service providers, competitive operators and mobile operators that have either 40Gbps, 100Gbps or both wavelength types installed in their networks, or that plan to install by next year (2013). 

The operators surveyed, from all the major regions, account for over a quarter (28%) of worldwide telecom carrier revenue and capital expenditure.

The study's findings include:

• A strong preference by the carriers for 100Gbps transport in both Brownfield and Greenfield installations. Carriers will use 40 and 100Gbps to the same degree in existing Brownfield networks while favouring 100Gbps for new, Greenfield builds. 

• The reasons to deploy 40Gbps and 100Gbps optical transport equipment include lowering the cost per bit, taking advantage of the superior dispersion performance of coherent optics, and lowering incremental common equipment costs due to the increased spectral efficiency.

• Most respondents indicate 40Gbps is only a short-term solution and will move the majority of installations to 100Gbps once those products become widely available.

• Non-coherent 100Gbps is not yet viewed as an important technology.

• Colourless and directionless ROADMs and Optical Transport Network (OTN) switching are important components of Greenfield builds; gridless and contentionless ROADMs are much less so.

Q&A with Andrew Schmitt

Q.  A key finding is that 40Gbps and 100Gbps are equally favoured for Brownfield routes. And is it correct that Brownfield refers to existing routes carrying 10Gbps and maybe 40Gbps wavelengths while Greenfield involves new 100Gbps wavelengths? What is it about Brownfield that 40Gbps and 100Gbps have equal footing? Equally, for Greenfield, is the thinking: "If we are deploying a new lit fibre, we might as well start with the newest and fastest"?

A: The assumptions on Brownfield versus Greenfield are correct, the definitions in the survey and the report are more detailed but that is right. 

It is more an issue that they [carriers] are building with 40Gbps now but will transition to 100Gbps where it can be used. Where it can't be used they stick with 40Gbps. There are many reasons why 100Gbps may not work in existing networks.

Q: Another finding is that 40Gbps is seen as a short-term solution. What is short term? And will that also be true for the metro or does metro have its own dynamic?

A: We didn't test timing explicitly for Greenfield versus Brownfield networks. It [40Gbps] doesn't necessarily peak, it is just not growing at the same rate as 100Gbps. And 40Gbps has even less value in the metro than in the core, particularly in Greenfield builds. With Greenfield 100Gbps combined with soft-decision forward error correction (SD-FEC), it is almost as good as 40Gbps.

Q: The study found that non-coherent 100Gbps isn't yet viewed as an important technology. Why do you think that is so? And what is your take on the non-coherent 100Gbps opportunity?

A: The jury is still out.

The large customers I spoke with haven't looked at it and therefore can't form an opinion. A lot of promises and marketing at this point but that doesn't mean it won't work. Module vendors are pretty excited about it and they aren't stupid.

Q: You say colourless and directionless is seen as important ROADM attributes, gridless and contentionless much less so. If operators are building 100Gbps Greenfield overlays, is not gridless a must to future-proof the network investment?

A: The gridless requirement is completely overblown and folks positioning it as a requirement today haven't done the work to understand the issues trying to use it today. This survey was even more negative than I expected.