Long run for short wavelengths?
The wait for a short-wavelength Fast Ethernet specification should end this quarter. The Telecommunications Industry Association's (TIA's) FO2.2 committee meets in Savannah, GA, Jan. 24-28 to address what it hopes are final comments on the second ballot of the draft it issued last July. If all the i's can be dotted and the t's crossed, and the next ballot is approved over the 30-day voting period, TIA/EIA-785 will make the use of 850-nm optics a standard alternative for Fast Ethernet applications.
The new standard may further open the door to fiber in the enterprise that Gigabit Ethernet has finally unlocked. According to Dan Silver, chairman of the TIA's Fiber Optic LAN Section and a marketing manager at 3M, the use of short-wavelength devices both decreases overall network costs and offers a clear migration from 10Base-FX Ethernet all the way to Gigabit Ethernet over multimode cabling. The hope is that the lower costs associated with 850-nm devices and the ease of upgrade to Gigabit Ethernet will appeal to end users looking to install Fast Ethernet networks with an eye toward a higher-speed future.
Certainly fiber in the enterprise could use all the push it can get. The cost/benefit tradeoff for fiber optics versus copper has always been clear: Fiber costs more for customers to install, but they get better performance in terms of bandwidth capacity and transmission distance. Unfortunately, this balance doesn't hold in the normal way with Ethernet. Both Ethernet and its Fast Ethernet sibling cap their bandwidth requirements at levels well within the reach of copper. Thus, fiber's only advantage for these applications lies in transmission distance--and the comparative lack of users who have been willing to pay the cost premium for long-distance 10-Mbit/sec transmission indicates how appealing this advantage has proven.
Still, when the IEEE established Fast Ethernet, the transmission-distance advantage of fiber was considered important enough to retain. Unfortunately, boosting transmission speed by 10 while retaining the 2-km distance attainable with the fiber-based 10Base-FX Ethernet standard required a change in optics to more high-powered lasers. Thus, the 100Base-FX Fast Ethernet standard included a requirement for 1300-nm lasers, instead of the 850-nm devices used in Ethernet. The distance stayed the same, but the cost went up, which naturally, didn't boost fiber's popularity with the end user.
Now that Gigabit Ethernet has arrived, the usual cost/benefit dynamics may again be in play. If forward-thinking network managers currently contemplating Fast Ethernet are uncertain whether their installed Category 5 copper will accommodate gigabit speeds in the future, fiber could prove an attractive alternative--provided the cost can be reduced to within the range of Category 6 or 7 copper networks.
The switch to 850-nm optics promises such cost reductions. Still, questions remain about the standard's effectiveness as a market booster. First, the amount of savings offered by a switch to shorter wavelengths remains elusive to quantify, says Silver. The number of features available also may affect cost. One big variable here is the potential provision of auto-negotiation between the new Fast Ethernet systems and existing Ethernet devices. Now that devices operating at both speeds will be sharing the same wavelength, Fast Ethernet hubs, for example, may have to negotiate speeds with legacy terminals operating via 10Base-FX. The question here, according to Silver, is whether system vendors think there's enough of an installed base of fiber-optic Ethernet devices to warrant the expense of implementing auto-negotiation in their new equipment.
The first products out of the chute will likely be media converters. Transition Networks has already announced both a media converter and a network interface card (NIC) that conforms to 100Base-SX. For the media-converter companies, auto-negotiation is a big plus, because it will enable end users to keep their old copper-based 10/100 NICs and still run fiber at least to the desktop outlet.
The fiber community has long touted the technology's ability to "futureproof" enterprise networks. We'll see if the new short-wavelength Fast Ethernet standard will finally get end users to listen.
Stephen M. Hardy
Editorial Director and Associate Publisher