A Peek at Packet-Optical Transport Evolution
100G Commercial Rollouts Begin
In the year since POTE last convened, 100Gbit/s (100G) Ethernet has become standardized by the Institute of Electrical and Electronics Engineers Inc. (IEEE) and momentum has increased. There are currently two suppliers shipping 100G systems for revenue, and between them they report more than two dozen commercial wins. Nearly every other major equipment supplier we've spoken with is planning to launch their 100G systems in either 2011 or 2012.
We believe such moves will be necessary to remain competitive in optical transport. In a Light Reading survey of 232 operators worldwide conducted last August, 53 percent of operators reported that 100G will impact their capex budgets in the 2011-2012 timeframe.
It's not all rosy for 100G, however. With the line side of 100G finally solidified, the industry has turned its attention to the client side. Here, the industry does not agree. One option uses 4x25Gbit/s lanes to get to 100Gbit/s of capacity (LR4) in the client-side module. The other uses 10x10Gbit/s lanes to get to 100G. The LR4 option has been standardized by the IEEE, but the LR10 version has some strong support -- including Google (Nasdaq: GOOG).
Another issue just beginning to surface is: what lies beyond 100G? Some companies are favoring 400G, while others prefer 1 Tbit/s. Work can't begin in earnest until the industry settles on a winner here.
OTN Switching Fills the Core
A year ago, Optical Transport Network (OTN) switching was talked about as an element of core packet-optical transport. Now, OTN switching appears to be the central requirement. According to some suppliers, OTN can satisfy both the optical and packet sides of packet-optical transport. New additions to the OTN, including ODU0, ODUflex, and ODUflex with hitless re-resizing, enable the switching of packets without "packet switching."
We know from our surveys that operators globally are very interested in OTN switching and the next-generation iterations. However, operators must make decisions on the roles of Layer 1 and Layer 2 in their cores before they deploy. "How much packet functionality should I put in my transport element?" is a very active question among operators.
Next-Gen ROADMs Gather Steam
The need for colorless, directionless, contentionless, and gridless reconfigurable optical add-drop multiplexers (ROADMs) was the central message of the keynote presentation from Verizon Communications Inc. (NYSE: VZ) at last year's POTE. For us, it marked our introduction to the topic. Since POTE 2010, we've conducted a significant amount of research on these ROADM functions -- which we place under the umbrella of next-gen ROADMs -- and have found that interest in next-gen ROADMs goes well beyond Verizon and well beyond North America as well.
One important point is that next-gen ROADMs can be broken into their various parts and deployed separately -- for example, colorless and directionless, but not contentionless; or all of CDC, but not gridless. Verizon has sets its sights on all of the above, but many other operators are still sorting out what's "necessary" from what's "nice to have." This year, we'll debate the pros and cons of each of the functions, as well as get an industry update on where the functions are in terms of commercial rollout.
Registration is still open for Packet-Optical Transport Evolution, taking place on May 18, 2011, at the Marriott Marquis in New York. Click here to register.
— Sterling Perrin, Senior Analyst, Heavy Reading