Iíd be somewhat leery of growth predictions from semiconductor research companies, particularly coming off a very bad year. Research companies only know one form of growth curve and thatís a hockey stick. ďISuppli projects that the industry will enjoy seven sequential quarters of growth before a seasonal downturn occurs in the first quarter of 2011, the firm said. ď This is August 2010. 2009 saw little or no growth. Predicting a downturn in 1st quarter 2011 isnít much of a prediction. If this is true, then this kind of predicts a double dip, doesnít it? I did a fast survey of historic growth predications for semiconductors in 2008(On the Web a prediction never dies. Isnít it great?) and some predictors(will remain nameless) were still predicating 7-8% growth for 2009 even as late as 3rd quarter 2008. Then you see the revised stats coming out in 2009. Predicting what will happen in 2009 in 2009, what kind of predicting is that?. What would have been really valuable if someone came out early in 2008 and foresaw the iceberg 2009 was going to be. Thatís what we pay for, but we never get it. We all buy the research stats because we need them to make some kind of estimate of how our business will be, since it takes too much time to do the research ourselves. 2010 should be a good year, since being in August, we can see it. 2011 and a double dip, who knows, but if inventories are building then that is prediction of a slowing 2011. A better predictor might be unemployment stats and how many people are in Best Buy or Frys buying electronics and how much the retailers are discounting to sell. The recent article about Android phones overtaking iPhones in 2012 is very positive. Competition means more semis sold, which is good no matter how you slice it.
The reality is that the industry can grow 30% and no job will be created in the US. Andy Grove's article shows that a typical PhD designer creates 7 jobs in Asia. We can grow anyhow, but the jobs will not come in the US, but overseas. So, that is the difference with all these rosy numbers that make those looking for jobs cry 'my country, why not help me'
What does it mean for the jobs scenario? Does nearly 30% increase in the revenue would also lead to more jobs in this industry? I am not familiar with the basics of this industry because in for example software tech companies the expected increase in supply already creates a lot of vacancies.
According to the article, some of this revenue growth is due to price inflation, but by how much, Dylan? Any unit sales figures?
As for the rest, nobody should be surprised at the uncertainty of the forecasts. The semiconductor industry is not hermetically sealed from the rest of the economy. In the same way as some economists are predicting continued growth of the American/EU/World economy and some are predicting a double dip, the outlook is also uncertain in the semiconductor industry. There are more optimistic voices than pessimistic ones though these days. Let's hope they are right.
Suddenly memory seems to be driving the semi's business and that makes me wonder, Is this growth fueled by core demand or some filling of backlogs?
I don't think we had anything interesting other than iPad this year and those big upcoming 3D TV's don't use memory like PC market.
Blog Doing Math in FPGAs Tom Burke 2 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...