LONDON RF power semiconductors for the wireless infrastructure sector may not be the best opportunity for companies developing and making devices using Gallium Nitride (GaN) process technology, a researcher at ABI (Scottsdale, Ariz.) suggests.
Apart from some military applications and microwave communications, most of the interest in GaN is centered on mobile wireless infrastructure and WiMAX. But according to ABI Research director Lance Wilson, the economics of this price-sensitive sector may mean that it is the wrong playing field for GaN.
"Aside from the newness of the technology, device cost is a huge negative for GaN when compared with Si LDMOS in conventional amplifier circuits," said Wilson.
In time, he suggests, incremental improvement in GaN pricing will overcome this problem. But Wilson says "GaN device pricing on a one-to-one basis will never approach Si LDMOS for mobile infrastructure. It is simply a more expensive process technology."
Even if superior performance is achieved, it will probably be only incremental, suggests Wilson. "In mobile wireless infrastructure, the two technologies are more or less equal in price and performance. GaN will capture some portion of the RF power amplifier business for mobile wireless infrastructure over the next few years, but certainly not all of it."
Wilson also maintains that there will be too many GaN participants for this market segment, and at least half will drop out willingly or be forced to do so.
However, he stresses that the process technology still has a bright future. Wilson suggests at frequencies above 4 GHz, beyond the performance range of Si LDMOS, GaN may dominate practically all of the high-power markets.
He adds Eudyna Devices and Toshiba have wisely targeted the microwave (>4 GHz) markets for much of their participation with GaN, and they will reap considerable benefits from doing so.