8/7/2007 -- So, what's it going to be? 40 Gbps (40G) has long been touted as the next-gen big pipe of choice, at least by many hardware vendors. Recently, however, a much bigger pipe -- 100 Gbps (100G) -- has started to make waves.

At this point, analysts say, vendors are already starting to push for bigger and bigger pipes -- even if customers aren't yet ready.

"[A] number of high-profile optical transport players are starting to talk about the nuts and bolts of delivering 40G. They're [also] starting to wax philosophical on 100G, and they're highlighting the necessary steps to get from 10G to 40G -- and beyond," wrote Jason Marcheck, a principal analyst with consultancy Current Analysis. "Similarly, standards bodies are getting serious about nailing down timelines and outlining technical requirements for various 40G and 100G implementations."

Vendors such as Nortel Networks, Ciena, Alcatel-Lucent, Nokia Siemens Networks and Infinera have all articulated either 40G or 100G (or both, in many cases) strategies.

"[W]hile 40G capabilities have existed in some vendor offerings since the early part of this decade, many are now making a concerted push to make 40G as compatible with existing 10G infrastructure as possible -- thereby minimizing the cost of implementing 40G wavelengths," Marcheck pointed out. "[M]any of the vendors that took the lead on 40G development in the early 2000s are now taking cues from large carriers like AT&T and Verizon and pushing toward having 100G products ready for commercial deployment by 2010. However, while several vendors are expounding their views on 100G, Alcatel-Lucent and Nokia Siemens Networks have clearly sought to become the frontrunners in the '100G stakes' by publicly talking up their product development activities, and even detailing early trial activities -- both in the lab and in field tests."

Even though Marcheck and other industry watchers anticipate a steady ramp-up to 40G, they also expect that 10 Gbps (10G) port sales will continue to dominate for several years.

"[E]xpect the spoils in the 40G market to go to the vendors that can best articulate a compelling 10G investment protection story for carriers looking to incrementally ramp up 40G," he said.

But 40G's ultimate triumph seems to have the stamp of a fait accompli, Marcheck said, in part because it's started to garner support from prominent standards bodies, such as the IEEE.

"The fact that the IEEE has decided to work on a 40GbE standard is a clear indication that some IEEE members have the clear intention of pursuing 40GbE product development; most would not do so if the market wasn't sending clear signs that 40GbE products will sell," he pointed out.

Marcheck, for his part, believes that both 40G and 100G will emerge as viable standards -- in part, he said, because 40G is already a fait accompli, while 100G represents the next logical step in the ever-expanding big pipe-stakes.

"[A] mélange of demand drivers, physical deployment considerations and standards development agendas should create a situation where both 40G and 100G solutions will have a role to play as optical networks continue to evolve over the next five years and beyond," Marcheck concluded. "[W]hile 40G is still considered an expensive commercial technology, 100G is still considered a largely experimental technology. This means that even when 100G becomes possible to implement in commercial networks, it will still take years beyond that point until 100G becomes viable from a cost perspective. Finally, while giants like AT&T and Verizon are pushing for 100G development, many smaller Tier-1/2/3 carriers will not have 100G in their network plans for years. For most of these carriers 10G will remain the preferred technology of choice for the time being, with the most cost effective higher capacity option -- again, most likely 40G -- representing the next logical step in the process." --Stephen Swoyer