Wanted: Applications to absorb fiber glut
This month's most shocking networking news, for its sheer magnitude, might well be the US$19 billion loss Nortel Networks Corp. reported for a single quarter.
Nortel President and Chief Executive Officer (CEO) John Roth blames the problems on the cutbacks service providers are making in their plans to build extensive fiber networks.
"We expect capital spending will only occur to remove network bottlenecks and for new profitable revenue opportunities," he says. Meaning there is enough backbone bandwidth built already.
There were similar signs from other companies.
If you look at Level 3's explanation of its financial situation last week you find that the company is laying off 1,400 of its 5,900 employees and cutting back on its projected capital spending through 2003. Why? Because many of its customers for high bandwidth connections are on shaky ground themselves, and Level 3 expects that will cut into their revenue base by perhaps 20 percent.
Tellabs will sell less of its optical gear than anticipated, and according to Merrill Lynch, here is why: "Service providers are temporarily able to meet increased customer demand for bandwidth by reallocating capacity within their networks and are only buying equipment to meet the immediate needs of their customers."
Optical gear maker Ciena, while expected to do well against competitors, nevertheless will experience less than the anticipated revenue, according to Merrill Lynch. The financial firm says it expects Ciena customers to cancel or delay orders, "especially in the U.S., where significant excess capacity exists. Once Ciena works through some of their recent contract wins we expect the company to experience some of the same problems that are plaguing the other optical system suppliers, namely customers canceling or delaying orders."
It is abundantly clear that service providers have built surplus bandwidth in optical core networks and that what is wanting is traffic to fill it. Pervasive high-bandwidth access networks made of fiber, DSL (digital subscriber line) wireless -- whatever -- are needed to feed this cavernous core. They are being built, but it will take time, especially now that competitive providers that were building such networks are fewer and fewer and less and less well funded.
Equally lacking are the killer applications that will drive demand for all this bandwidth. And that is what is really needed. Wall Street pulled the plug on competitive access providers because they weren't coming through with revenue, and lack of compelling applications was at least part of the problem.
If service providers had a compelling high-bandwidth service to sell, people would buy it. Fast Internet access wasn't it. Video conferencing and distance learning don't seem to be it. But, eventually, something will be.
It may take a few years and new wave of bright young engineers who take this gargantuan fiber infrastructure as a given to come up with that compelling use.