What The Chip Wars Taught Intel’s Andy Grove
Bloodied by the worst slump in its history, U.S. semiconductor makers, says Grove, will have to be “agile giants,” and pugnacious protectors of their technology.
November 1 1987 by Chief Executive
Last April, Andrew S Grove, 52, became CEO of the Santa Clara-based world leader in microprocessors.
An influential, outspoken industry figure, he gives new meaning to the phrase “hands-on manager.” Having emigrated from Hungary in 1956, Grove (ne AndrasGrof), worked his way through New York City’s College as a waiter, then researcher, and ultimately pursued a doctorate in chemical engineering at The University of California at Berkeley. It was there that he met Gordon Moore, then research director at Fairchild Semiconductor. When Moore and Robert Noyce left Fairchild to form Intel, they took young Grove along.
Grove is a relentless operating man, something he wants to get away from.
This might be difficult since his reputation preceeds him: He is well-known for his nationally syndicated. business-advice column. (Putnam published a selection of his `Dear Abby of the workplace” columns in One-on-One with Andy Grove.)
Having emerged from a $255 million bloodbath over two years, Intel has slimmed operations, closed obsolete plants and shifted its focus to more specialized technologies. Although IBM has reduced its stake in Intel, it remains a big customer (the successor to the Intel 80286 the 80386-is at the heart of IBM’s Personal System 2.)
In contrast with his precedessor, Grove has taken a pugnacious line in litigating against unauthorized “improvements” in company chip design-a practice historically discouraged but tolerated in Silicon Valley. No more Mr. Nice Guy. Expect more courtroom brawls over intellectual property. In talking about the industry’s future over a luncheon salad in his office cubicle (his is only slightly bigger than the rest at Intel’s Lakeside Drive headquarters), Grove doesn’t disguise his contempt for free trade. “That guy is a jerk,” he once said of Nobel economist and free trader Milton Friedman. He sees Sematech (the industry consortium set up with defense department help to improve semiconductor manufacturing skills), as a useful, if limited, weapon in the continuing struggle with Japan’s electronics giants. Others are not so sure.
Imagine your newspaper column receives the following letter: Dear Dr. Grove, I am a recently appointed chief executive whose high-tech company emerged from a ravaging recession. Even though our products are now very much in demand, we are nonetheless troubled by vigorous foreign competition. What should my priorities be?
Sounds familiar. Too bad I don’t correspond with myself! I can answer relatively easily, if somewhat facetiously, with the phrase, ‘make hay while the sun shines.’ By that I mean, take advantage of the situation your products are in (demand) while you slim the company down. Since the question obviously pertains to Intel, I would add that as long as there is a good fighting spirit within the company, use the opportunity to prepare for the next slump in demand. Figure out what you want the company to be like when it happens, and rush madly toward it while conditions are comparatively benign. What if you don’t know what the conditions are going to be like? You reinforce a degree of agility in your managers’ thinking. Fight off the notion that conditions which recently have turned better, will stay better. More importantly, encourage a frame of mind within the company that combats anything which may push you back. Use the money you make today to reinvest-not in those products which are winning for you now, but in those which will diminish your dependence on them. Above all, diversify. (I don’t mean buy hotel chains.) Prepare for future turmoil by developing product offshoots and broadening your range.
What lies beyond the next slump? What are you racing towards?
There has been strong growth in the personal computer industry since 1985. This won’t last forever so it is important for us to prepare for changes by supplying more products for use in personal computers-in addition to the central processing unit such as the 286 and 386. Besides, the computer that everyone sees the terminal-on-a-desk computers-are proliferating behind the skins of other machines-cars, refridgerators and automatic teller machines, for example. As the market becomes more sophisticated, designers want more variety in their products. As a result, they are taking our standard computer family and mixing it with customized chips. Because of this emergence, we can expand the number of places where we can stick our machines. The money we make in supplying CPUs is being used to develop a very expensive ASIC (Application Specific Integrated Circuits), capability. We started doing this in late 1985 long before the horrible $150 -million quarter loss, because we saw that this is a necessity. One application of this is our work with a manufacturer of VCRs. We have not previously participated in this market because it would have been too costly for us to develop a circuit for them because VCRs use simple components. Our capabilities lie with more complex circuits. We are now seeing the two markets converge. The VCRs are getting smarter and require our type of sophisticated circuitry.
How would you characterize the industry slump from which you emerged?
By far the most vicious. The money lost by the U.S. semiconductor industry was 10 times that of the previous slump. I estimate that the American industry lost $2 billion in this recession.
That agrees with Dataquest estimates.
It’s a back-of-the-envelope estimate of Japanese losses. My first estimate was also $2 billion.
Dataquest figures the same. Maybe they used your envelope.
I have a new number. An economist, specializing in our industry came by here and said that a second, more careful look, revealed that whereas the U.S. loss is correct, it appears that the Japanese loss is closer to $4.5 billion in the same two years while the Koreans lost $1 billion. Compare this to what happened in the previous down cycle. Industry losses are easily 10 times as large. Why? It’s the presence of the Japanese and the Koreans. After the slowdown in 1981 and ’82, the personal computer industry came roaring back, soaking up microprocessors and memory capacity, ultimately saving us. During the recession of 1974-75, the Japanese were not a major factor. In 198586 they were. Their persistent market share, as opposed to profit orientation, exacerbated the damage. This last recession was different, in that it was a bare-knuckle fight with the Japanese.
What lessons have you drawn from this experience?
I did not learn anything I didn’t already know. I know the need to be ready and able to cope with wrenching change. It has almost made me paranoid in terms of driving us toward ASIC capability. The fact that we had to shut down obsolete facilities and build new ones was not a revelation, but I probably would not have done it at the same rate. Necessity sped things up. We are much more disciplined about what commitments we make with our customers than we were in 1983 and 1984.
Has your new role as CEO changed you?
It’s urged me to clean up my act and get myself out of operating tasks. I grew up as an operating manager and liked it. However, there is a huge need at Intel for a `foreign minister,’ one who manages external relationships. Our job is to design, build and sell things. Now we have different entanglements. We sell items to people we compete against and litigate against. A perfect example is our largest customer in Japan, NEC-the same company whose semiconductor division is in legal with us. On the other hand, NEC is among our top five customers worldwide. We support them with great belief and diligence. Due to these frictions, we did not do such a wonderful job supporting them as a customer during the last recession. We couldn’t separate in our minds their role as a customer from their role as a semiconductor dealer. This kind of turmoil requires a CEO who’s more of a foreign minister. I think it is a good time for me to become one. Besides, it is healthy for all concerned for me to leave my operating role. There are better people around to take over.
Last February, you said the semiconductor industry was, “past the point where the industry can take care of itself. This isn’t the 11th hour, it is 12:01.” What does the clock read now?
Since I made that remark, government has stepped up its efforts-in a very aggressive fashion-to stop the Japanese juggernaut. It was unprecedented particularly for this administration, to come through with the punitive tariffs. It had an impact on Japanese industries and on the behavior of Japanese competitors.
But didn’t tariff sanctions have the perverse effect of increasing semiconductor prices to your customers, the computer makers?
That did happen, but that was before the sanctions. Bringing chips in by the suitcase full and other subterfuges took place in the absence of strong governmental action in Japan. That caused the U.S. government to finally say enough.
But the computer industry was forced to buy more expensive chips. Isn’t the whole purpose of the thriving semiconductor industry to aid the computer industry?
Absolutely. But the American computer industry is not at a disadvantage because of the suitcase problem. The biggest danger the computer industry faces is the disappearance of its principal domestic supplier. In fact, the U.S. computers would be in mortal danger if American semiconductors were destroyed.
Let me take you back a step. What happened in television before technology changed black and white TV? The U.S. was dominant. The Japanese industry, then very small, dumped its products to gain market share. Sanctions or penalties were upheld but not imposed. When the industry moved to color TV, the Japanese were in parity and the process continued. American companies were driven into the losing proposition of shutting down, one-by-one, their factories and selling them to the Japanese at bargain basement prices. The Japanese established production capability with the surviving U.S. producer, Zenith. The next product in that line was the VCR. There wasn’t an American company producing in that industry. In less than 20 years and three technological revolutions from black and white TVs to color TVs to VCRs, the Japanese achieved full dominance.
Meanwhile, the industry has grown from a billion dollar industry to a $80 to $90 billion industry. Now substitute say D-RAMS, EPROMS and microprocessors and you see what’s upon us.
If the Japanese lost $4.5 billion to the U.S’s. S2 billion, isn’t that enough to dissuade anyone from this kind of tactic?
That’s why you need the government to carry the other half of the burden. Government’s duty is to make sure this kind of purgatory action is not practiced by a collection of foreign competitors, arguably aided, coordinated, and encouraged by government policy. Fairness is not the consideration here, it’s self interest. It is not in the self interest of American semiconductor or computer companies, or the U.S. economy, for this to continue.
Let me give you a statistic. For every dollar of goods consumed in the U.S., American industry historically produced a dollar of product. Today, for every dollar that we consume, we produce 80 cents. That’s horrendous. When it happens in the most vital element of the economy, which is the information processing segment of electronics it’s the equivalent of having our domestic airlines in somebody else’s hands. O.K., we can live without a VCR-making capability, but our information processing infrastructure is going to be vital to the survival of the economy.
Yet the same argument-”Help us, we’re a strategic industry”-put forward by steel companies 10 years ago had the result of preserving sclerotic steel companies. Is that what you want for semiconductors?
First, on behalf of that industry, there is a resurgence of steel.
Yes, led by second-tier firms, none of which were protected.
Oh no. They are protected somewhat. The fact that the steel industry may not have done the right thing doesn’t mean that government thinking was wrong. Maybe what was wrong was the steel industry.
Assuming semiconductor chief executives-unlike the steel chiefs of 15 years ago-are far more prescient, knowledgable and aggressive, what incentives exist to prod them into taking corrective action if they sit behind protective barriers?
Don’t think that the incentive to survive has been removed by this miniscule action of government. We are talking about $300 million worth of sanctions, which is nothing more than a clearing of the throat in an electronics trade deficit of $20 billion. The government has merely signalled its intention of aligning itself with survivors of the semiconductor shakeout. A temporary modification of Japanese business practices is all that has happened. No big deal. That won’t ensure our survival. Don’t think for a moment that American semiconductor manufacturers are a bunch of patsies. We compete with each other as aggressively as the Japanese. There is one important difference: We have shareholders. We have a lot of motivation to drive costs down, export to new markets and come up with clever devices for new applications. That’s the only way we are going to stay ahead of everybody from Motorola to NEC. The only people that ever get paid by the customer are those who are ahead of the pack.
Don’t you think this industry is beginning to mature?
Absolutely not. It’s heading for another upswing after a major realignment. People are getting out of businesses which they only had half a reason to be in in the first place. We’re no longer in magnetic bubbles or E2 memories. It requires too much money to do a good job, so we decided to take the money and put it into other things. Intel is a perfect example. The Fairchilds and Mosteks and other troubled companies are the visible part of the shakeout. The invisible part is where companies are narrowing their competitive focus. In the 1970s, we all thought we could do everything equally well. The industry has become very brutal, very heavily financed by every government-ours being the last to get involved. The German and Dutch governments are pouring billions of dollars into project Mega. The Korean government is using our money to fund their semiconductor industry with which to attack us.
And you think Sematech will make a difference?
Sure. It will make a difference. This problem doesn’t have a single answer. We can not afford to pass up efforts that will give us a boost. The Sematech consortium is a way for the industry to support its suppliers by defining clear targets to shoot after, so that we do not, in turn, get too dependent on overseas vendors for manufacturing tools. By getting together to come up with prototypes, members can get a degree of security for development and direction.
Assuming all you highly competitive people can agree on something as complex as that.
It may not happen, it may not work. It recognizes that times have gotten permanently more competitive. For example, we can’t give technology away for a pittance anymore. In the past, we let technology flow where it might, and by running a little faster, one could be successful. The rules are different today. We are not going to compete by trying each other out for new equipment. Manufacturers deal because that is the way the Japanese will win. Sematech will help us collaborate on that. We also compete on intellectual property more. Why? Because the cost to develop intellectual value is skyrocketing. You can’t treat a $100 million piece of property like a $100,000 deal. The relative contribution of well- defined intellectual content as compared to the physical material is really going out of whack. The recognition this shifting of value leads to is a more protective attitude. It’s true of everyone including Texas Instruments, not just Intel. A new pattern is emerging. People who were on the receiving end-getting million dollar technologies for a thousand dollars-want to continue to receive those technologies. The people on the creative end may lose $900,000 or a million dollars, but I am sure as hell not going to forfeit $10 million. Until a new order of valuation for these things takes hold, there is going to be a lot of friction in this industry.
Who will dominate the submicron area, the U.S. or Japan?
Probably neither-in a technological sense. Although we will fight tooth-and-nail over it, it won’t be more than a portion of the battle. The submicron area gets you to the point of putting multimillions of transistors on a chip. What you put on that chip and how you put it on is going to be at least as big a competitive weapon as etching fine lines. The investment in developing the computer to design that chip is almost equal to what it took to develop the process. As we develop sub-microns the balance may very well shift to the computer-aided design, checking, and test-generation capability. There may be a number of companies that will be able to make submicron circuits, but they may not know how to market them. The winning company will have the submicron capability in its computer tools and the market knowledge to utilize it .