A Bit Of This, A Byte of That

The problem now is to read the tea leaves for anticipated recovery in 2002 correctly. Worst-case analyses suggest that almost no vertical market in telecommunications will experience true upturns within the coming calendar year. If this turns out to be true, more layoffs might be in store by mid-year. Maybe it's time to follow the lead of one engineer at October's Communications Design Conference, who thought it might be best to put digital filtering DSP expertise to the service of 3D protein-folding applications in biotechnology, since communications would be shot to hell on a semi-permanent basis. Such a prognostication might not be too extreme, given the continued grim results at many of the larger telecom carriers.

Somehow, though, I have a feeling there will be some signs of life present in certain sub-markets by spring. Top management at the leading semiconductor companies and OEMs will be playing a balancing act between spreading bets around to avoid chasing a single chimera, and funding new research realms adequately, in order to introduce products for new markets in a timely manner.

Before we drill down to specific sub-markets, a necessary warning is in order: Market rallies will come and go during the course of the winter, but they should not be misinterpreted as true signs of life at the OEMs. Given the condition of telecom carrier and enterprise spending, the earliest we can hope to see the first signs of a sustained recovery would be mid to late spring. There's no call for being needlessly pessimistic, but we all need to be realists, as well.

It's likely that the first signs of sustained recovery will come in enterprise applications. LAN technology may be mature, but expansion of Gigabit Ethernet and 10 Gigabit Ethernet business for enterprise backbones will represent the first blossoms in spring, as will inter-system interconnect, one reason why Communication Systems Design's departing Editor-in-Chief, Rob Keenan, has been telling you to pay attention to your interconnect technologies.

The next likely market to rally will be the front-end systems that interface with server farms, such as caching systems and web switches. These will take care of the temporary server bottlenecks that enterprises and ASPs will experience in early 2002. The next step out will be edge systems for WAN access, which is why most router companies are spending more time these days talking about edge routers, rather than core systems.

In the metro, the incumbent carrier will be the sole customer in several applications. That will favor next generation SONET systems or systems that look like SONET. Resilient Packet Ring, for example, could have an edge if the RPR Alliance gets its act together. Ethernet may have the advantage of familiarity, as well as orders from the likes of Telseon and Yipes, but if incumbents control the market, Ethernet still must prove that its resiliency is as good as SONET. Similar trends will be true in metro wireless backbones: everyone talks about all-IP for the future packet 3G network, but people want resiliency behind their packets.

Core routing and optical cross-connect markets will take the longest time to recover. The folks working in dispersion compensation for 40-Gb networks could have a bright future, if they can hold out for several flat quarters. But the long-distance carriers will be moribund through most of 2002. That could mean a tough few months ahead for optical cross connect (OXC) and terabit router vendors. But then, we're all facing some tough months, aren't we?