Silicon Insider: Semiconductor Sense

ByCommentary by MICHAEL S. MALONE

July 14, 2005 -- -- Hey, let's talk about the most boring technology subject I know: semiconductor equipment manufacturing.

Wait … don't go away. Because semiconductor equipment is also the most important tech industry, especially if you want to see, and especially invest in, the future.

Semiconductor equipment manufacturing (henceforth SEM) has always been not only the most ignored part of the electronics world -- and the least understood -- but also the most benighted. It sits at the absolute beginning of the digital food chain, as the companies that make the equipment that makes the chips that make the devices that make the hardware that runs the software and games and makes the phone calls for the world.

Unfortunately, rather than giving SEM companies priority over the rest of the digital universe, they more closely resemble a position equivalent to the poor guy at the end of a crack-the-whip game. SEMs have always been the tail of the tech dog, lashed back and forth, the last into good times and first into bad.

Amazingly enough, it used to be even worse. For the first 25 years of the industry, up until about 1980, SEM companies not only suffered all of these systemic problems, but they also had a nasty habit of either growing to a certain size and dying; or having one great product, then messing up the follow-up product … and dying. It was an industry composed mostly of little companies, struggling to get through the next downturn, and perpetually at the mercy of their semiconductor industry customers -- the nastiest and most demanding clients imaginable (how would you like to have been a supplier to Andy Grove and Intel?)

All of this changed with the arrival of Jim Morgan as CEO of Applied Materials. Unlike most of the other great executives in high-tech history, Morgan's genius was for consistency: he turned Applied Materials into the most reliable of companies, always putting out solid products generation after generation, never taking great risks, but perpetually pushing forward the state of the art.

In the process, Morgan, who just retired, not only built Applied into a multibillion-dollar company, in the process stabilizing both the SEM industry and increasing the reliability of every chip built in the world, but he also turned Applied Materials into best corporate citizen in Silicon Valley. And he did it in the most mercurial and unforgiving of industries -- for which he deserves to be in the pantheon of great business executives of the 20th century.

Morgan managed to solve the internal puzzles of the semiconductor equipment world, but no one can cure the larger forces that whipsaw the SEM. It goes with the business. Too bad for them, but good news for the rest of us. That's because, for anyone watching closely, SEMs are canaries in the coal mine, the ultimate early indicator of good times or bad times ahead in high tech.

Unfortunately, few people ever do watch closely -- despite the advantages that vigilance can confer. And there's a good reason for that: the fabrication of semiconductor chips is unbelievably complicated, closer to processing chemicals or cracking petroleum than, say, assembling automobiles. And the equipment that does this work is even more arcane than that. In other words, unless you happen to have the letters MSEE or MSME after your name, just thinking about this stuff will make your head explode.

Trust me. A decade ago I wrote a book about microprocessors, one of the most sophisticated of semiconductor chips. In that book I wrote a long chapter about how chips are made -- and apparently, I did a pretty good job, as that narrative is still taught in some college engineering intro classes around the country. But writing that chapter was one of the worst experiences of my professional career; there were days when my head hurt so bad I thought I was having a cerebral hemorrhage. And even now I can't read more than a few paragraphs of that chapter without my eyes glazing over and finding myself inexplicably in the kitchen cracking open a beer.

And folks, keep in mind, I get paid for writing about this stuff. If I find myself in the middle of writing a story about epitaxy and wafer sort and ion implantation and seriously consider moving to another state, changing my name, and taking a job at Wal-Mart, just imagine what it would be like for you to voluntarily develop an expertise in this field.

That's why you don't read many stories about the SEM industry either. Most technology reporters come into the business from the world of games or software or e-business. Many don't understand how a PC works. So they don't even like writing about the semiconductor industry, much less the equipment that builds the stuff. With the exception of Dean Takahashi at the San Jose Mercury-News, I think most tech reporters would rather gnaw off their writing arms than cover the semiconductor equipment beat.

The irony is that SEM people are among the sanest and smartest people in the high-tech world. They have to be to survive. And when you talk with them, as I have quite a bit in the last few months, you gain a powerful perspective on the digital world that you can't find anywhere else: if chip orders are trending up or down, who is investing in the next generation of chips, where the next breakthroughs are coming from, etc. They see the winners and the losers, the booms and the busts long before the rest of us.

This week, the semiconductor equipment industry is having its big annual Silicon Valley conference, Semicon West. And, as it has for the last three decades, the sponsoring trade association, SEMI, used the occasion to make near-term economic predictions for the industry. And, as it usually has for the last three decades, most of the press ignored the story. You shouldn't.

Here's why: after polling its members around the world, SEMI predicted that the industry would slump 12 percent to $32.6 billion this year. Don't trust the precision of those numbers, but only notice the direction -- especially since the industry grew by a whopping 67 percent last year. The prediction is also considerably worse than the 6 percent drop for the year that SEMI predicted last December.

In other words, having come out of the 2001 crash hot and fast last year, the SEM business has stumbled. This suggests that, the bubble aside, the SEM industry is still operating on the traditional, and predictable, four-year boom-bust cycle. It also suggests that, as a leading indicator, this year's slump in SEM presages a similar slump next year in chips and, about six months after that, in all of high tech.

The good news is that these same surveys suggest that this downturn will be much shorter and far less devastating than the crash of 2001. In fact, the SEM world anticipates a return to modest (8 percent) growth in 2006, faster growth (10 percent) in 2007, and the industry should be rolling along smartly again (14 percent) in 2008. I don't really believe that last number: if the market really takes off again, it could three times that, but you get the idea. Better yet, most of the SEM companies, having ramped staffing up slowly the last couple years, believe they will get through this year without any serious lay-offs.

What does all of this mean? That after a disastrous start, this decade is likely to see nearly continuous growth in the tech sector in the second half. There will likely be a shallow and short downturn in tech late next year, but the industry will come out of it quickly and be headed toward a major boom that will peak in 2009. Interestingly, two new forces will help drive that next boom --a score of new chip plants in China, and the rise of nanotechnology.

All of this means that, if you are going to play the market in tech stocks, you've only got a few more months of good times before the indexes flip over. If you are still looking for a job after the last crash, you are running out of time. If you are planning to start a new tech company, now is the time to do it, when you can still get the venture capital money, and with the prospect of a year or so of reduced competition to get your product or service ready for the market. And if you are working for a large tech company, now might be the time to get out of high-risk ventures and find a nice safe slot on the organization chart.

And everybody should get ready for one helluva ride from 2007 to 2009. If you are going to get rich in tech, that will be the time. So be ready to make your move.

There. Now see: semiconductor equipment isn't that boring after all, is it?

This work is the opinion of the columnist and in no way reflects the opinion of ABC News.

Michael S. Malone, once called “the Boswell of Silicon Valley,” most recently was editor at large of Forbes ASAP magazine. He has covered Silicon Valley and high-tech for more than 20 years, beginning with the San Jose Mercury-News as the nation's first daily high-tech reporter. His articles and editorials have appeared in such publications as The Wall Street Journal, The Economist and Fortune, and for two years he was a columnist for The New York Times. He has hosted two national PBS shows: "Malone," a half-hour interview program that ran for nine years, and in 2001, a 16-part interview series called "Betting It All: The Entrepreneurs." Malone is best known as the author of a dozen books. His latest book, a collection of his best newspaper and magazine writings, is called "The Valley of Heart's Delight."

ABC News Live

ABC News Live

24/7 coverage of breaking news and live events