During a lifetime in the RF/microwave industry, one gets to see a great deal of change. Take silicon-bipolar-transistor amplifiers: When they first started to replace vacuum-tube power amplifiers some three decades ago, tubes were replaced in high-frequency designs (sometimes begrudgingly at first).

Then came the "GaAs explosion" in the 1980s, with both military and commercial companies pushing to discover the potential of this then novel semiconductor material. On the military side, Raytheon Co. devoted much time to the development of discrete GaAs power transistors and GaAs monolithic-microwave-integrated-circuit (MMIC) amplifiers, intended for such applications as phased-array radar systems.

On the commercial side, companies like Pacific Monolithics of Sunnyvale, CA made enormous contributions to the development of impedance-matched GaAs MMIC amplifiers for use in point-to-point radios, as well as in the rapidly growing satellite-communications (satcom) markets of the 1980s. Firms such as California Eastern Labs/NEC pioneered the development and application of discrete GaAs transistors for both power and low-noise commercial amplifier applications.

Of course, GaAs technology received more than a little push forward from the US Defense Advanced Research Projects Agency (DARPA; www.darpa.mil) and the large amounts of cash injected into companies working on GaAs devices as part of the agency's GaAs MIMIC program (about 1988 to 1999). Although most of the DARPA investments were meant to aid the needs of military applications, it was also understood that the technology's benefits would eventually reach the commercial sector.

DARPA invested so heavily in GaAs because it appeared to be "the solution" for high-frequency, solid-state integrated-circuit (IC) design and for high-power discrete transistors. That is, until the potential benefits of GaN transistors and ICs became more apparent. What DARPA did for GaAs, it hopes to do now for GaN, as part of its Nitride Electronic NeXt-Generation Technology (NEXT) program.

The DARPA NEXT program's stated goal is to achieve faster transistors using GaN technology. Current technologies, such as silicon complementary-metal-oxide-semiconductor (CMOS) and silicon germanium (SiGe) heterojunction bipolar transistors (HBTs), can reach cutoff frequencies to 200 GHz and beyond, but only at very low voltages. The NEXT program is attempting to develop a technology with wider voltage swings, taking advantage of the high breakdown-voltage levels of GaN semiconductor materials. (For a more detailed comparison of semiconductor processes, please see the "Industry Insight" feature).

In many ways, it could be argued that DARPA's investment in GaAs made the "cellular communications revolution" possible. If DARPA's NEXT program helps GaN technology develop as GaAs did, just imagine the long-term commercial benefits.