Getting interoperability in place, and then standards will speed along the process. IMO, ADSL will be a low margin/high volume play.
There are the obvious players, and I would look for consolidation in the industry - like Paradyne goes public in a couple of weeks (they have the Bremer based MVL and Superline tech) then maybe early next year, if enough LECs are using their equipment, they get bought out by LU or CSCO (who BTW like their Netspeed acquisition, but seem to have made a mistake with Degaz) or other telco system integrators. Maybe PAIR (who is working on tech that will allow 2G DLCs to handle ADSL, has their low cost, low power ASIC Falcon DMT chip, and their Avidia high density access concentrator) buys AFCI, which then makes them an even more compelling consolidation target with all the usual suspects. TI's C6X is still pretty power consumptive and dissipative, as well as large and pricey, but maybe they get that silicon act together. In the meantime, Paradyne uses their C5X. And MOT is making some familiar grunts again. And Nortel owns Elastic Network's compelling Etherloop tech. Even Globespan's recent DMT silicon purportedly has some impressive specs (as well it should, since they pioneered CAP tech) and will be going public in the next month or so. ORCT/Fujitsu is definitely worth watching. BRCM is kinda skipping over ADSL and primarily working on VDSL (along with their DOCSIS silicon) - which probably won't kick in big anywhere soon (US Worst is trialing it in Phoenix), but they KNOW silicon and MIGHT be able to bring something to the ADSL table last minute. Also, CNXT has been doing a lot of work on DSL silicon (though Dwight has lost some bright engineering talent to the two Henrys)
IMO, it's really up for grabs, and I don't see any big-wave "pure play" favorite at this time, though RBAK from the IPO and aftermarket has been a GREAT winner for me recently.
Personally, I think broadband fixed wireless is a compelling longer (AND shorter) term play for high margin, bottlenecked commercial broadband access. |