IBM and the Future of Microelectronics

With the proliferation of low cost end user devices that can connect wirelessly to the Internet, the need for a major increase in microchips is just beginning.  When IBM chose to build its next generation 300 mm plant, dubbed Fab 2000 in Fishkill NY, many thought it was the end of the plant here in Vermont.  It probably would have been if the need for microprocessors to power games, global positioning systems, etc., did not give the older 200 mm plants a new lease on life.  In other words, despite a political climate that influenced IBM’s decision to build its next generation microelectronics plant in Fishkill, rather than here in Vermont, where they already had a plant that was at the cutting edge of the industry from a technology perspective, the plant here in Vermont survived.

Now we have a political climate that is even worse for IBM deciding to remain here in Vermont.  At the same time, market conditions are pointing towards the need to drastically increase the capacity of semiconductor manufacturing.  Not only are the number of devices using a microchip proliferating, but the market for such devices is growing exponentially.  In addition to a the growth of demand in the U.S., low cost electronic devices that can connect wirelessly to the Internet are becoming popular rapidly in formerly developing countries as they see a rise in their middle class.  This rise is demand comes at a time when the traditional cost effective methods of meeting such demands are running into snags.  This methods were to make smaller chips so that more chips can be created out of a single silicon wafer and to use bigger wafers.  The industry is now getting to the point where it is increasingly difficult to make the chips smaller, or the wafers bigger.  The result will be a need to increase production capacity by building more sites that manufacture chips and to expand sites already in operation.

The other factor involved here is that these end user devices do not require as much power and traditional computers, but they do require more functionality.  This favors manufacturers who specialize in sophistication, rather than low costs.  Such a trend may very well reverse to process of seeking to outsource manufacturing solely based on low costs.  In addition to sophistication in regards to functionality, there is also the efficiency of the process being used to manufacture semiconductor devices.  A higher yield in chips/wafer can more than offset higher labor costs.  In bother of these cases, IBM’s technology leadership positions it well to thrive.

Besides the proliferation of devices that require low powered microchips, there is the race to build more powerful chips without creating a bigger “foot print”.  The ability of getting more power out of smaller chips is also staring to run into hurtles.  The next step is to stack chips and create a tower-like three dimensional structure resembling a sky scraper.  Here again, IBM is in on the cutting edge of this trend.

What does all this have to do with Vermont?  We were once home to the plant, which IBM referred to as its “Early Production Line”.  It was at the cutting edge of new technology and would have been a logical place to have Fab 2000 if the political climate was not so hostile here.  We lost that opportunity to Fishkill and only succeeded in keeping the current scaled down version of the plant because an increase in market demand made it not feasible to close it down.  Now there are new market opportunities coming our way in the microelectronics field.  Will we take advantage of these opportunities by encouraging IB to meet the coming demand increase by expanding the capacity of its plant here?  Are our political leaders even aware of the possibilities?