I would point that a sequential fall of about 1.0 percent for Q4 in comparison is normal for the semiconductor industry.
I make the average of Q4/Q3 fall over the last 10 years 1.4 percent, according to WSTS data, so a sequential fall of only 0.7 percent in better than average.
Hi Peter - the latest Cowan LRA Model's expectation for 4Q/3Q sequential sales growth (as calculated from the WSTS's recently posted October actual sales of $24.873 billion) is minus 1.6 percent which is slightly more negative than your calculated last 10-year historical average of minus 1.4 percent.
Hi Brian - the latest run of the Cowan LRA forecasting model incorporating the WSTS's October actual sales number of $24.873 billion yielded an updated 2012 sales growth expectation of minus 2.3 percent which is in excellent agreement with IHS-iSuppli's latest 2012 yr-o-yr sales growth update of minus 2.3 percent. The model's 2012 sales forecast estimate came in at $292.758 billion compared to the WSTS's 2011 reported sales of $299.521 billion. Remember that IHS-iSuppli's global semiconductor sales tracking process is different than the WSTS's and this accounts for their higher prediction for 2012 of $310 billion. Also note that the latest WSTS sales and sales growth forecast expectations (per the WSTS's recently published 2012 Autumn forecast update) are $289.936 billion and minus 3.2 percent, respectively.
Peter - WHOOPs. The 4Q/3Q sequential sales should be PLUS 1.6 percent, that is, a sequential INCREASE, NOT minus as stated above.
Therefore, the latest Cowan LRA Model run for 4Q/3Q's sequential sales growth forecast estimate is, in fact, more "bullish" than iSuppli's prediction of plus 0.7 percent sequential sales growth and slightly more positive than your stated last 10-year average of 1.4 percent. Sorry. Mike C.
In this single Intel notebook example across three process nodes.
iCore 7/5/3 total quad & duals production:
Arrandale 188,630,122 units
Sandy Bridge 156,025,668 units
Ivy Bridge 138,566,718 units