In my first EE Times column, I waxed eloquent about the outlook for communications ICs. Well, the severity of the recent meltdown came as a surprise. To mend the error of my ways, I have thought hard about how financial and technology analysts missed this turn and what will lead to a much-anticipated recovery.

I built my optimism on what pundits call "the insatiable demand for broadband." We should know better. After all, in the closely associated Internet sector, the post-mortem taught us that e-commerce is not a business unto itself, but rather just a way to sell products and services. So, why did we think bandwidth was an end unto itself?

Even the most learned observers were caught up in bandwidth momentum thinking without considering the fundamentals in the business and consumer markets. With the collapse of new business formation in the business-to-business and business-to-consumers arenas, carriers and service providers could easily plan for capital expenditure cuts. Some suggest that carriers have slowed spending to maintain the bottom line in a slowing economy, but I do not buy that argument fully. This just would not be profit-maximizing behavior if bandwidth demand were truly insatiable.

The consumer sector is more difficult to figure out. Broadband connections-DSL and cable-represent less than 10 percent of all connectivity. If only the installed base of computer users shifted en-masse to broadband, then the communications systems and semiconductor industries would experience phenomenal growth.

Perhaps the problem lies in a killer app gap. The ways consumers use their computers and the Internet have gotten us this far. But evolutionary or revolutionary applications will be needed to fuel the next wave. Meanwhile, a chicken-and-egg problem holds back consumer bandwidth demand. Consumers would benefit from a richer multimedia experience on the Net. However, Web sites cannot shift to such content until consumers have fatter pipes; and so it goes.

A month ago, most analysts pronounced that a recovery would arrive by midyear without stating what would fuel it. Now, analysts say the recovery will happen early-2002. I am seeking the killer apps that will drive the recovery in the consumer and business sectors. Learning at least from one mistake, metropolitan optical access technology will not fuel the recovery. It will be applications that demand this amount of bandwidth.

Jeremey Donovan is a principal analyst for communications at Gartner Dataquest (jeremey.donovan@gartner.com).