40 Gbit/s: Ready for Prime Time?
ANAHEIM, Calif. -- OFC 2002 -- Members of the 40G Collaborative hosted a breakfast forum yesterday morning here at the Optical Fiber Communication Conference and Exhibit (OFC) to dispel what they see as a myth about the emerging market for 40-Gbit/s optical telecom technology (see 40G Collaborative Formed ).
Despite rumblings to the contrary, this group of component and system vendors says 40-Gbit/s technology is ready for deployment. All that is needed are carriers willing to deploy it.
“Not only are components available today, they are available from multiple vendors,” said panelist Aileen Sansone, manager of technology development for JDS Uniphase Inc. (Nasdaq: JDSU; Toronto: JDU).
Indeed, several companies made 40-Gbit/s product announcements at the show this week:
- Lucent, Corning Team on 40-Gig
- Inphi Intros Mux/Demux and TIA Chips
- Centellax Intros 40-Gig Amps
- Velocium Debuts 10-Gig, 40-Gig Gizmos
- JDSU Demos Everything
- AMCC Unveils Chips Galore
- Spirent Demos 40-Gbit Test Solution
- Transpectrum Demos Transceiver
Despite the progress, analysts remain skeptical. “I don’t think that 40 Gbit/s is ready for prime time,” says analyst Rick Schafer of CIBC World Markets. “[Vendors] may be able to build some basic components, but they can’t do it in volume. So I would argue that they aren’t there yet.”
Panelists yesterday conceded that much 40-Gbit/s technology works only in certain environments under certain conditions. For example, PMD is not a major problem at 40-Gbit/s rates when it is running over newer, more modern fiber.
“Forty-Gig running over fiber installed today or tomorrow will not have PMD issues. And it can still transmit signals up to 1,500 or 2,000 kilometers at 40 Gbit/s," said panelist Claudio Mazzali, strategic alliances manager for Corning Inc. (NYSE: GLW). He explained that the fiber quality and the age of the fiber play critical roles in how much dispersion is present in a system.
Unfortunately, large amounts of older fiber will be subject to problems, particularly in long-haul facilities where distance also increases CD and PMD effects. Even though there are new devices that can minimize these issues, their use adds more cost to the network.
“It comes down to economics,” said panelist Mark Barratt, VP of business development and product planning for LaserComm Inc. “I watched the migration from 2.5 Gbit/s to 10 Gbit/s, and the fact is, carriers won’t be interested until they see at least a 50 percent cost reduction per bit for 40 Gbit/s.”
Schafer says economics make the debate over whether or not the technology is ready irrelevant. “It’s almost a moot point to argue,” he said. “Even if the technology was perfect and available, the carrier market is in a holding pattern. We are all frozen, waiting for the capex spigot to turn on again.”
The panelists agreed that the market won’t take off in volume for at least another year. Phil Francisco, panel moderator and VP of marketing for PhotonEx Corp., a long-haul transport startup, said he expects gradual deployments at first, as carriers upgrade their most bandwidth-starved routes with 40-Gbit/s technology.
“Some 40-Gbit/s systems are already in trials now,” he said. “But you likely won’t see anything like a normal adoption pattern until 2003. Some of the new deployments will be 40 Gbit/s, and some will be 10 Gbit/s, depending on the network.”
— Marguerite Reardon, Senior Editor, Light Reading
http://www.lightreading.com For more information on OFC 2002, please visit: www.nottheofc.com