Four hot topics for 2008

by Stephen Hardy

OFC/NFOEC, the optical communications industry�s most important event, moves this year from March to February 24�28. That means that this year�s Executive Forum, co-presented by the Optical Society of America and Lightwave and co-located with OFC/NFOEC in San Diego, will be in February as well, on February 24�25. As usual, attendees will hear C-level executives from both well-established and up-and-coming component, subsystem, and systems suppliers offer viewpoints on the latest technology and market trends. (You can find out more about the Executive Forum at http://www.ofcnfoec.org/special_events/index.aspx.)

One new wrinkle this year will be a panel entitled �Four Hot Topics.� As its name implies, the panel will feature executives who will discuss four areas the forum committee members believe will influence the optical communications space during 2008. In the interests of full disclosure, I will admit that I�m on the forum committee. In fact, I was one of two committee members assigned responsibility for this panel, so in large part the topics represent what I think will be hot in 2008.

Let me tell you about them and what I expect you�ll hear our panelists say.

The first, high-speed communications, is the proverbial �no brainer� in terms of inclusion. On the 40-Gbit/sec side, it appears that most carrier applications will involve some form of alternative modulation format. The choice of format is beginning to coalesce around two alternatives: DQPSK in Japan (which best matches the optical properties of the predominant fiber type deployed there) and DPSK everywhere else (because the technology is cheaper than DQPSK). But other alternatives continue to be explored�perhaps with an eye toward giving them a trial run at 40 Gbits/sec in order to ready them for 100-Gbit/sec applications.

The problem is that no one�s really sure what the first generation of 100-Gbit/sec technology will look like for carrier applications. We can feel confident that data centers, where 100-Gigabit Ethernet will likely see its initial deployment, will see some sort of parallel, rather than serial, transmission implementation. That�s why parallel optics will be our panel�s second hot topic. Even if we agree that 100-Gbit/sec will first appear on the market in a parallel format, we don�t yet know what combination of data rates and transmission streams�4�25, 5�20, 10�10, or some other combination�will prove most efficient and economical. I�ll be interested in what our panelists have to say on this topic.

But, of course, parallel optics has already found a home in applications that require much less than 100 Gbits/sec. The current data center high-water mark, 10 Gbits/sec, has already offered opportunities to parallel optics vendors, as applications ranging from 10GBase-LX4 transceivers and QSFP devices for InfiniBand attest. I expect our panelists will highlight what the future holds for these applications, as well as new areas of opportunity.

And one of these new areas will undoubtedly be our third hot topic, consumer electronics. Anyone who has purchased an HDTV is now familiar with HDMI cables, which could just as easily be optical as coaxial or any other copper-based medium. Cell phones are more frequently using short optical links to drive displays. Meanwhile, both the IEEE 1394 FireWire and USB specifications have received data rate upgrades into the gigabits, which may open the door for more optical ports on home electronics devices and personal computers. I wonder if this may finally open the door a bit for plastic and glass optical fiber in the home�and I wonder if companies whose experience revolves mainly around the telco or data center environments can successfully cash in.

Many of these advancements and opportunities I�ve discussed here so far will be enabled by new developments in our fourth hot topic area, wafer-level integration. Whether the avenue is silicon photonics, photonic integrated circuits, or some other approach, several companies have found new ways to squeeze more functions into a smaller space. The first trick, of course, is to be able to produce these multifunctional devices economically in high volumes. The second trick is to base these devices on a technology or process platform that provides room for even tighter integration. We�ll undoubtedly hear arguments for and against the various approaches our panelists espouse.

These four areas don�t represent the only topics or trends that will shape 2008. Undoubtedly optical technology will continue to march deeper into the access networks of telcos and MSOs alike. Carriers will continue to look for ways to efficiently handle packet-based traffic via reconfigurable links. But these four topics are a good place to start a discussion of what 2008 will bring�and I hope you�ll join that discussion next month.

Stephen M. Hardy
Editorial Director & Associate Publisher
stephenh@pennwell.com

More in High-Speed Networks