Larry: Gallium Arsenide is like injection molded ceramic motor car's engines, they are the materials of the future. They were both the future fifteen years ago and they are still the future today. Now, mind you, in a $100,000,000,000/year market, GaAs companies (add to your Triquint, Anadigics and Vittesse) can grow very rapidly and more rapidly then the semiconductor industry in general. However, GaAs will not challenge the dominance of silicon until it provides twice the performance at half the cost. There are a number of intrinsic reasons why this is not on the horizon, not the least of it is the still poor quality of GaAs wafers relative to silicon (thus lower yields) and more importantly a basic factor of at least five in the cost of such wafers (and this is the theoretical factor, not the one currently in the market). There a number of growing niche market where GaAs has established itself, but their problems is that everytime they establish a new frontier of performance (and in many respect GaAs is better in performance than Si) Si creeps back into the game for these applications and beat GaAs on yield and thus cost. Intrinsically at the present time they have the field of optoelectronics and very high frequency communication to themselves. The former is guaranteed for good, since Si simply cannot provide light emitting devices, the latter is subject, however, to the creeping advancement in Si technology.
Zeev |
