One hope of those talking about climate change in Copenhagen is that solar panels will eventually be cheap enough to compete unsubsidised with fossil fuels. As the semiconductor industry was inspired by Moore's law to make computer chips tiny, powerful and ubiquitous, it's thought that technology and competition might work a similar magic for photovoltaic cells. Indeed, semiconductor equipment maker Applied Materials has started pushing into thin film solar equipment. And last week Taiwan Semiconductor Manufacturing Company, the world's biggest contract chipmaker, took a 20 per cent stake in solar company Motech Industries for $193m. Both suggest that industrial cross fertilisation might be about to begin.
Yet the similarities between chips and panels are only superficial. Both involve silicon and electricity. They also share some processing technology. But solar is a much simpler process, with very high production volumes. A typical plant will churn out 3,000 silicon wafers every hour, with each wafer forming one solar cell. A reasonably sized semiconductor foundry, by comparison, might produce only 200,000 wafers in a year. It is a business of minute precision, given that one wafer can yield 100,000 chips.
Routes to improved efficiency also differ. In chips, it is largely a question of better design. For solar panels, it most often happens by procuring raw materials more cheaply. One reason why margins at solar panel producers are so poor is that many locked in silicon supplies when prices were high last year.