This article is very confusing. Why don't they just publish the revenue numbers for each segment instead of this confusing array of percentages and multiples?
2010 177.6M, 2011 457.3M, 2.6 times growth ok.
2014 $1.4B, 45% Compounded from 2011 to 2014, ok
Established MEMS 10% growth 2011-2012, ok
New MEMS triple digits growth? Need to calculate
New CE&Mobile MEMS 31% to 62% growth from 2012-14
New CE&Mobile 39.5% share by 2014 so much less today.
If the whole market is growing at 45% and the 70% share is only growing at 10%, do we presume that New CE&Mobile is growing at 122%? If 80% share is growing at 10% then New CE&Mobilie growing at 167% Why don't they just state the number?
I agree with VincePG that this article would have been easier to follow if it presented a graph with year-on-year growth. The value of the article would have been heightened if the descriptions detailed all of the factors contributing to this large growth. How did they earn what they earned?
In addition, were there any other analysts backing the report? What was the report based on? Also, the author could write about the significance of IHS iSuppli as a market research firm. We have to be careful of reports claiming such high growth when there are environmental, legal, and other unforeseen factors at play.
Finally, having lists to describe all of the types of devices would be easier to scan than having them scattered in sentences throughout. However, if these projections are indeed accurate and realistic, then I can't wait to see the actual growth rates!
Don't forget that iSuppli and similar research firms are in business to sell subscribers and others the full report. Many times the 'teaser' data is presented to lure in more buyers. If you really want to know, buy the report. But remember, a forecast is just a forecast, and is good only at the time it was made. With a base data point and Excel, you can do as well :=). The key value of research data clearly lies in what is historically true (e.g. past performance) and in trends that can be verified using alternative sources. Quarterly updates are needed to see if everything is going according to original forecasts (they don't) and whether the trend is still correct.
When developing a model (linear/nonlinear)for demand forecast based on historical data, we must assume the calibrated parameters will not be changed in the future and this is impossible in a changing world !
Blog Doing Math in FPGAs Tom Burke 24 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 ...