That's quite alright. I just like to chat as an occasional distraction. And, if you happen to be interested in the IC Industry:
Like most modern inventions, the IC industry has quickly proliferated from the genesis-technology nations of America and Europe to eastern developed and developing nations such as Japan, Korea, and Taiwan. Although the US can still be very much considered a leader in the IC industry (the US makes up 4 of the top 10 IC companies)1, this is quickly changing. Not only are companies like Samsung and Toshiba outpacing US companies like Texas Instruments and Qualcomm, but US companies are increasingly setting up foundries and offices oversees. Some companies, such as Qualcomm and AMD are selling off their foundries altogether.2 While the US is still competing overall in the IC industry only 2 of the top 10 foundries (GlobalFoundries and IBM) are US-based.3 This shift to outsourcing creates several problems:
1) A decrease in the number of available jobs, in general, is bad for the US economy,
2) Innovation occurs where the most work is being done; while IC innovation from any country is an overall benefit, the US is set to potentially lose out on future technology by allowing other countries to outpace them in the IC industry,
3) A decrease in jobs creates a decrease in demand for education for those jobs, causing universities to become outdated in those fields (this same problem occurred in the power industry and there is now a great demand for power engineers and few universities to train them).
There seem to be two possible approaches, each with convenient examples within the industry. There is the fabless model (eg AMD, which spun off its fab to GlobalFoundries), in which IC companies have essentially given up fabbing their own devices. Then there are companies such as IBM and TI, which have retained their fab divisions. The fabless model creates two problems. Obviously, a company cannot help to bring fabs back to the US if they don't own any fabs. More importantly, however, these companies seem to be shooting themselves in the foot. Not only are they forcing themselves to rely on outside companies for their products, they are literally paying the R&D budgets of future competitors. Started as a dedicated foundry, TSMC has now moved into the design field, such as LED lighting.4 It is not unreasonable to think that in the future TSMC could become a direct competitor to companies like AMD with the added advantage of full vertical integration.
How to get IC jobs, and especially foundries, back in the US is certainly a prime example of the chicken and egg question. Companies will not start foundries in the US if there is no one to work in them and no one will train, or offer training, to work in a foundry with no job prospects. This is distinctly different from the current issue in the power field; while there was also a significant decrease in power engineer graduation rates the last couple decades, people will notice when their electricity stops working. A lack of foundries is a more subtle problem. Government initiatives (such as the one that created ECE571) would certainly help. While industrial foundry jobs are lacking in the US, there are still an overwhelming number of IC design and simulation (which, unlike fabbing, is an ever-increasing field in the US) jobs. There is no reason students couldn't supplement their design course with fabrication courses, so that they are prepared to work in foundries should those jobs become available. Government initiatives would not only help preparing engineers, but could also be used to motivate foundries to move back to the States. Examples of this already include renewable and nuclear energy and the auto industry (if the government can afford $1 billion to salvage an industry that was fully responsible for its own collapse, surely it can afford to bring a vibrant, healthy industry back to the states).
It should be noted that the foundry business in the US is certainly not dead, nor are the aspirations of inspired students interested in fabrication, but not in leaving the US. GlobalFoundries, based in California and the 3rd largest foundry, has a brand new 300mm fab in New York5 and Intel is planning on a new 300mm fab in Arizona within the next few years.6
need 1 more for 3 day/anon/WTA/full press game
' bet, 3 day deadlines, anonymous, full press and winner takes all. Reply or send me a message for the password if you're interested. Thanks!
)
)
)
)
)