I do not believe we are not even close to the "period of long, slow decline" stage with the semiconductor businesses.
They allowed what could be done to control the timing rather than what should be done.
For years, the "SIA" road map (now "ITRS") had a 3 year cycle for technology nodes. This began compressing in the late 90's, (unofficially at first, as individual companies demanded faster delivery, then officially) to 2 year cycles.
In my opinion, this did not allow the companies to profit from their previous investment before moving on the next. Once this race started, all companies had to join it or be left behind.
This resulted in product being delivered with capabilities the consumer was not yet interested in using, because the applications had not had time to mature, at prices that were falling because of the aggressive, node pursuing, over-capacity.
This slowdown will slowly reverse as applications mature and the semiconductor companies develop some restraint. My guess (and it is purely a guess) is that when the semi's are fully recovered, the industry will be very close to where a unmodified 3 year cycle would have placed it. However, without the profits that would have resulted from staying with that 3 year cycle.
Just because something can be done, does not mean it should be done. An oil company COULD increase production, lower prices at the pump and increase market share in the short term and make a lot of money. However, they would deplete their reserves and be out of the oil business in a few years if they did so. They SHOULD pump at a rate that will maximize the return over the life of the asset.
The Semiconductor companies forgot this simple rules for a few years as they unrealistically dreamed of getting an overwhelming advantage on their competition.
They will be much better from now on....<ggg>.
Ira |
