We just received a press release from National Semiconductor Corp. (NSC) noting that it is their 50th anniversary, click here. While NSC is not the oldest semiconductor company, it is among that senior group, of course. Other companies in the multi-decade club include Analog Devices, Linear Technology, Maxim Integrated Products, Texas Instruments, Fairchild, International Rectifier, just to name a few (yes, I know I have left some vendors out; this is not a definitive, complete list by any means, so please don't get upset).
Historically, 50 years is a very long life for any commercial enterprise, but I am not sure to what extent we can or should apply historical trends to our industry. There are a lot of ways to look at this, some complementary, some contradictory, some fascinating. We're an industry that has been driven by innovation and constant re-invention, by road maps, by disruptive technology events, and by the need for major, major investment in R&D and production. Not only has “solid state” and ICs advanced basic electronics and physics, but it has driven a world of ultra-precise metrology and ultrapure materials.
But is reaching 50 years the end of the beginning, or the beginning of the end? Obviously, I don't know, nor does anyone. On one side, our industry continues to spend significant time, money, and energy to push the limits of Moore's conjecture (sorry, I refuse to deem it a “law”, despite its apparent validity), even in these recessional times. On another side, we see a relatively our linear road map, which is sometimes a bad thing because at some point, you come to the end of that road, and then what? And on yet another side, there are the mind-boggling costs involved in pushing along that road map, which means you need enormous volumes to pay the bills, and maybe those volumes can't be sustained.
Then I see the tremendous sums that are being invested in “biotech”, a broad, catch-all term for genetic research, tailored pharmaceuticals, medical devices, and related disciplines, and the huge “life sciences” institutes that have been built at MIT and Harvard. Maybe these are hints that the era of ICs and semiconductors is coming to an end as a synonym for “high tech”, and there's a new high tech coming on strong?
Personally, I don't think that our industry is entering a period of old age and decline. Certainly it is changing, just as the relentless push for ever-bigger and faster CPUs has been transformed into a multicore architecture at the high end, and an incredible diversity of microcontrollers at the other end. As we approach the inevitable end of progress largely derived by shrinking the process (it can't go on forever), we will have different approaches and kinds of innovation, with powerful software that can do some pretty amazing, if not scary, stuff–have you seen the latest in computer-generated special effects, or CAD (computer-aided design) and simulation tools?
Let's check back in 50 years, to see what happened. Until then, take some time to look at any of the many good books, oral histories, and on-line “museums” on the history of the semiconductor industry. There are many good ones are out there, so I'll suggest “Spinoff” by Charles E. Sporck (Saranac Lake Publishing, 2001; ISBN 0970748108).
While the past is not a predictor of the future, it can still be interesting and enlightening, especially in our fast-moving industry where almost anything more than a few years old is considered “antique” and almost irrelevant at first glance.: ♦