MTT-S, from my own experience, is not necessarily just "a bunch of bald guys," but certainly engineers whose image of the world to come is tempered by experience. We assume a more youthful crowd will make its presence felt at two upstart conferences: the Penton Wireless Symposium/Portable by Design event (www.wirelessportable.com) in San Jose, Calif., next week; and the International Microwave Symposium in Boston on June 11 (www.ims2000.org).

The first conference will feature a lot of introductory and tutorial material on cellular and wireless LAN technology. But the exhibitor list is dominated by semiconductor makers like TI, Analog Devices, Motorola and Intersil-manufacturers who believe they can tune BiCMOS and CMOS processes for RF design.

Currently, gallium arsenide is still the technology of choice for RF, especially the antenna drivers in cellular handsets. GaAs is arguably the most efficient device at 1- and 2-watt power levels in the 900-MHz and 1.8- and 1.9-GHz cellular bands. And these power transistors are relatively cheap: less than $1 each in high volumes.

But GaAs has some disadvantages. The transistors are generally depletion-mode devices. Always on, they need a negative bias to turn them off, and that's not always easy in a battery-powered device like a cell phone. Plus, GaAs cannot be integrated with baseband processors in CMOS. Thus, upstarts at the Wireless Symposium are working on processes that may eventually integrate RF, intermediate-frequency and baseband functions to make a single-chip cell phone.

As for the International Microwave Symposium, a daylong workshop last year concluded that the single-chip cell phone is (at least in the short term) impractical for RF devices requiring complex modulation schemes like cellular handsets, but that new semiconductor technology could easily build an integrated RF front end for devices with limited range and power like Bluetooth.