Полная версия
Inside Intel
The same cannot be said for the third 10x force that faced Intel. Towards the end of the 1970s Japanese semiconductor makers started to pose a serious threat to the memory-chip business that had always been Intel’s cash cow. The issue was complicated by roller-coaster market conditions that prompted the Japanese firms to ‘dump’ their products on the American market, selling them not only below the prices they charged in their home market but also well below their manufacturing cost. But the underlying problem, which Grove and Moore refused to face up to, was that Japanese chip companies paid more attention to quality and spent more effort trying to perfect their manufacturing processes than Intel did.
The result was that the lead that Intel had created in the memory-chip business began to erode – and by the middle of the 1980s the company’s refusal to accept that it needed to go back to school was threatening its very survival. Only when Andy Grove and Gordon Moore asked themselves what they would do if they were a new management brought in to clear up the mess did the solution emerge. Intel pulled out of memory chips, savagely cut back its workforce, and refocused its firepower on microprocessors. This decision, analysed exhaustively in business-school case studies and magazine articles, has been hailed as one of the company’s finest moments. But the praise begs the question of whether Intel could have solved the problem earlier at lower cost. Had Grove responded to the 10x force of Japan more quickly, might Intel today be twice as big and profitable as it is?
In 1997 it is the Internet that Grove identifies as the most powerful 10x force facing Intel, along with the rest of the computer industry. So far, few fundamental changes to the company and its operations have seemed necessary. While Microsoft has made extensive changes to its applications products and operating systems, basing them on a new vision of working in which almost all information from daily calendars to reports is published and exchanged over computer networks, Intel’s response seems to have been more muted. The company has issued an extension to the instruction set of its processors, called MMX, which allows computers to process sound and pictures more efficiently. It has made some astute venture-capital investments in a number of the more interesting Internet startups. It has helped to push the PC industry towards building computers that cost less and are easier to install and maintain.
But Intel would look very different in a networked world in which individual users had less computing horsepower and fewer bloated software packages on their desks, relying instead on smaller, simpler and faster pieces of software downloaded across the Internet as needed. In such a world the extensive installed base of software that is compatible with Intel technical standards would be much less of an advantage. Yet the company’s response so far has been to do little more than point out, with some justification, all the shortcomings and flaws in this vision, and to pour cold water on the much-hyped Network Computer promoted by Oracle’s Larry Ellison. It is all too tempting to wonder whether the moment when Intel’s triumph appears sweetest might – like the moment when IBM’s mainframes seemed secure in their domination of the computer business – be the beginning of its downfall.
As it faces the challenges ahead, Intel has a number of strengths. Its management team, almost entirely developed internally, is extremely strong. Its corporate culture allows the company to set objectives, communicate them swiftly to its workforce, and make a good attempt at achieving them. Its compensation system, which rewards hard work and loyalty with stock options worth millions, but checks underperformance with regular reviews and ‘corrective action’ programmes for laggards, is highly successful in motivating Intel people to give their best. And its lack of hierarchy makes it easier to respond swiftly to change and to make rational decisions.
But the Intel that Andy Grove has created also has its weaknesses. The company has been plagued by arrogance since its earliest years. It has frequently taken a high-handed approach to its customers, and suffers from the ‘Not Invented Here’ syndrome as badly as many technology companies. Most alarmingly, the company has found it increasingly hard to accept outsiders into its senior ranks. Like transplanted organs, managers brought into Intel from outside have more often been rejected by the patient than absorbed.
These weaknesses are likely to come into renewed focus when Andy Grove departs from the scene. To many insiders, a post-Grove Intel is still unimaginable. After managing the company’s operations for two decades, and more recently guiding its strategy too, Grove has become almost synonymous with Intel. Yet he passed his sixty-second birthday before this book was published, and had a narrow escape from prostate cancer in 1996.
In theory, the succession is settled. Craig Barrett, Intel’s chief operating officer, was promoted to the company’s presidency in May 1997. He now officially handles the company’s day-to-day business and is ideally placed to succeed Grove on his retirement. But there must be a question about whether a less forceful, less driven personality than Grove will be able to lead the company with the same success.
Ultimately, the deciding issue will be people. And it is people, not technologies or strategies, who are the focus of this book. Its aim is to offer an account of Intel’s story as seen through the eyes of dozens of different employees, from the most junior to the most senior. The lives of these people don’t add up to a comprehensive history of the company. Since the company has always refused to cooperate with outsiders attempting to tell its story from an independent standpoint, that may have to wait for many years until the secrets of Intel’s current operations are no longer of commercial value. Instead, the intention here is to give a glimpse of life inside Intel – and in doing so to say something about one of the most extraordinary and most ruthlessly successful businesses in history.
PART I
INNOVATION
‘We are really the revolutionaries in the world today – not the kids with the long hair and beards who were wrecking the schools a few years ago.’
GORDON MOORE, Intel founder,
quoted in Fortune magazine
1
The Odds-On Favourite
YOU PROBABLY THINK you can skip this chapter.
The scene is already in your mind. It’s late at night, and the garage is entirely dark except for the pool of light cast on the workbench by a low-cost anglepoise lamp. The future billionaire is hunched over the computer, oblivious to the clutter of empty pizza boxes around him, absorbed in his work. His hair is unwashed, and he’s been wearing the same grimy T-shirt almost every day since he dropped out of college. He has few contacts and no backers. His only assets are his technical skills and the brilliant powers of persuasion and negotiation that will blossom over the years to come.
There’s probably no single company that conforms to every one of these stereotypes. But most of America’s most successful technology companies display at least some of them: modest beginnings, fighting against the odds, brilliant ideas that go against conventional wisdom, founders who are outsiders and have nothing to lose if they fail. Look at Steve Jobs and Apple, or Bill Gates and Microsoft. These are the models that we’ve come to think of as the ways to start a successful high-tech company.
The creation of Intel Corporation in 1968 was quite different.
Instead of being young and rebellious, its two founders were middle-aged and respectable. Instead of being poor and isolated, they were prosperous and known already as leading figures in their industry. Instead of labouring for months or even years to find a backer for their venture, they rounded up $2.3m of funding in an afternoon, on the basis of a couple of sheets of paper containing one of the sketchiest business plans ever financed.
The two most important words of the business plan were Robert Noyce. Forty years old, Noyce was the general manager of Fairchild Semiconductor, one of the most prominent businesses in the Bay Area to the south of San Francisco. But he was more than that: he was one of the creators of the integrated circuit.
To understand the significance of this, you have to remember that the earliest computers used vacuum tubes as the basic elements of their circuits. Vacuum tubes, working like small-sized light bulbs, were bulky and unreliable – and since they had to be heated before they could work properly, they were also gluttonous consumers of electricity. A large computer could easily be big enough to require its own little power station – and its vacuum tubes pumped out enough heat to turn a massive room into an oven.
The building-block of today’s electronics industry is a miniature switch that takes advantage of the fact that certain crystals, such as silicon, sometimes conduct electricity and sometimes don’t. This switch, dubbed a ‘transistor’, earned a Nobel Prize for the three physicists at Bell Labs who discovered it in 1948. The early transistors were smaller, and needed no heating element to make them work. Moreover, unlike a light bulb they didn’t need to be changed every so often. But they shared one of the drawbacks of the vacuum tube: to build a computer, you had to connect them one by one into electrical circuits.
Bob Noyce’s claim to fame was making it possible to put more than one transistor on to the same fragment of silicon. The circuits built using this technique became known as ‘integrated circuits’. Coincidentally, two different teams in different companies 2,000 miles apart conceived the integrated circuit almost simultaneously in 1959. The winner of the first integrated-circuit patent was Jack Kilby, an engineer at Texas Instruments. But it was Noyce and his colleagues at Fairchild Semiconductor who succeeded in turning the integrated circuit from a laboratory prototype into a product that could be mass produced in ever-increasing numbers and ever-falling prices – and Noyce who had made it possible for engineers to dream of myriad new products that had never before been possible.
Noyce did not fit the stereotype of the inventor. He was gregarious, charming, athletic and handsome. Brought up in Grinnell, a small town in Iowa where his father was a Congregational minister, he was a boy scout who went to Sunday school every week and graduated valedictorian at the local high school. His entry in the school’s year book described him as the Quiz Kid, ‘the guy who has the answers to all the questions’, who played in the school band, sang in its chorus, was a leading light of the Latin and science clubs, and acted in six plays. At college, he was the swimming team’s best diver, and took the lead in a radio soap opera. The sole cloud over his exemplary youth, which formed the centrepiece of a profile of Noyce that Tom Wolfe wrote for Esquire in 1983, was an incident at college when he and a fellow-student stole a twenty-five-pound suckling pig for a Hawaiian barbecue on campus. People took pigs seriously in the Iowa farm belt. But it was a sign of the charmed life that Noyce seemed to lead that not even an offence of this gravity could spoil his career. An admiring physics teacher talked the college authorities out of expelling him. Instead, he was sent off to work at an insurance company for a few months, and then after his graduation to the Massachusetts Institute of Technology, where he earned his PhD.
More than two decades later Bob Noyce had become general manager at Fairchild and chairman of the board of trustees of his old college in Iowa – but he was still everybody’s best friend. He had a way of looking at you that made it clear that he took what you said very, very seriously, and a way of talking in his gravelly, deep voice, usually waiting until everyone else had weighed in first, that made you take what he said more seriously still. He had what Wolfe described as the ‘halo effect’. (‘People who have it seem to know just what they are doing; they make you see their halo.’) He was not only a born leader; he was also an inspiration.
Gordon Moore, reporting to Noyce at Fairchild Semiconductor as head of research and development, was a very different personality. While Noyce was five foot eight and dark-haired, Moore was over six feet and balding. While Noyce was the life and soul of every party – drinking, singing, playing tricks and accepting every dare offered to him – Moore would sit with a few close friends, quietly talking around a table. His temper was a model of equanimity, and his two great passions were fishing and messing around in boats. He was born in the small coastal town of Pescadero, just thirty miles south of San Francisco, where his father was deputy sheriff of the county. If you bumped into him walking out of his local hardware store on a Saturday morning, which people often did, you’d find him in a pair of worn overalls, peering down at his solid work boots through his wire-framed glasses. You might easily have mistaken him for a modestly prosperous orchard farmer, out in his pick-up to buy something he needed to fix a leaking pipe or to put up a swing for the kids on the apple tree in the back yard.
But Moore was every bit as great an engineer as Noyce. His PhD, in chemical engineering, was from Caltech, the prestigious California Institute of Technology near Pasadena. He was the winner of a number of important patents. And he had an uncanny knack for solving technical problems. If you took a problem that looked as though it had five or six routes to a possible solution, most engineers would waste a lot of time exploring and then ruling out the ones that didn’t work. Not Moore: for some reason that neither he nor anyone else could explain, the one avenue of enquiry that he chose would often be the one that yielded the best results.
Moore was legendary for sitting down at the lunch table in the company cafeteria with a group of engineers who had been battling with a problem for several months, and suggesting, very quietly, that they might like to take a look at this or try that. More often than you would expect, they would go away and take his advice – and discover that fifteen minutes of Gordon Moore had brought them closer to a solution than months of unaided work. Moore was also a gifted listener and judge of character. Slow to anger, he would hear everyone out. Only in the presence of people who went on for too long, repeating an unnecessary point far beyond the point of boredom among others, would Moore’s self-control finally slip and his anger become visible.
When a local paper asked Moore in 1968 why he and Noyce had decided to set up a new company, his answer was that they wanted to experience once again the thrill of working in a small, fast-growing company. Not for the first time, he might have added. Noyce and he had been involved in startups twice already. In 1956 they had helped William Shockley, the leader of the team at Bell Laboratories that developed the transistor, to set up his own laboratory. A year later the pair had been among a group of eight who walked out on Shockley to start their own transistor company under the banner of Fairchild.
The mass defection from Shockley Laboratories had been provoked by the old man’s irascibility and paranoia, and his insistence on ignoring their advice in a technical dispute over which area of research they should concentrate on. To raise funding for the new company, the ‘Traitorous Eight’ struck a deal with Sherman Fairchild, an inventor on the East Coast whose father had been one of IBM’s earliest investors. Fairchild had advanced $1.5m to them in 1957, on the understanding that they would create a subsidiary of his Fairchild Camera and Instrument Corporation specializing in the manufacture of semiconductors. If the venture failed, Fairchild would pick up the tab. If it was a success, the company would have the right to buy them out for $300,000 apiece.
It was a success. Two years later Fairchild’s company exercised its option, and made them all rich men. The eight founders were left in place at Fairchild Semiconductor – but they no longer had any real control over the business, nor any great stake in its success. Noyce was promoted from general manager to group vice-president of the parent company, and was held in the highest respect by its owner; but everyone knew that the big decisions were always made by the accountants back in Syosset, New York.
For a while the problems remained invisible. As the 1960s wore on, however, members of the original team became bored and drifted out of the company to set up on their own. Fairchild Semiconductor became the electronics industry’s equivalent of a sycamore tree with its winged seeds: every season, seeds from Fairchild would spin away gently in the wind, land somewhere nearby and burst into growth as new saplings. Meanwhile, as Fairchild’s growth began to slow, the money men from New York began to come up with strange compensation schemes, in which the general manager of each part of the division was paid a bonus linked to the current profits of the business he was responsible for. This turned cooperation into rivalry, and made the managers reluctant to put into effect the technologies developed by Moore and his team. New technologies might make money in the long term, but not in the short – and in the fast-moving environment of Silicon Valley, who could tell where you’d be working by the time the payoff came?
The bell tolled for Fairchild Semiconductor in 1967, when Charlie Sporck, the company’s famed manufacturing genius, set up in competition at National Semiconductor, and hired away busloads of his former colleagues. It required no great insight to tell that Fairchild was irreparably injured, and that the best Noyce and Moore could do would be to slow, rather than staunch, the flow of blood. A year later the two men were the last of the eight founders left.
The decisive moment was a conversation between the two men on a sunny weekend afternoon, while Noyce was mowing his front lawn. Once the partnership was established, the next step was clear. Moore, as head of Fairchild Semiconductor’s research and development (R&D) department, would start work on the products the new business would develop. Noyce, who eleven years earlier had led the negotiations with Sherman Fairchild that created the company they were now planning to leave, would raise the money.
It took Noyce just one phone call. The deal with Fairchild a decade earlier had been brokered by an investment banker from New York called Arthur Rock. Rock had come out to California to see the Traitorous Eight; he had helped them to draw up a list of potential backers, and he had shopped their idea to thirty-five different companies before getting a yes from Sherman Fairchild. Since then, Rock’s life had become intertwined with the Fairchild founders. He had moved to California and set up a new investment bank of his own in San Francisco, specializing in the financing of new companies. (Today, this business is known as ‘venture capital’; the phrase was invented by Rock himself.) Rock felt bound by loyalty to Fairchild to avoid taking any active steps that would hasten the break-up of Fairchild Semiconductor. But he had already helped two of the founders of the business to set up on their own, and Bob Noyce was his best friend among the remaining six. The two men used to take hiking and camping holidays together.
‘Bob just called me on the phone,’ said Rock afterwards. ‘We’d been friends for a long time … Documents? There was practically nothing. Noyce’s reputation was good enough. We put out a page-and-a-half little circular, but I’d raised the money even before people saw it. If you tried to do it today, it would probably be a two-inch stack of papers. The lawyers wouldn’t let you raise money without telling people what the risks are.’
Rock made fifteen phone calls that afternoon, and received fifteen acceptances. His goal was not simply to raise the money for Noyce and Moore: with their track record and his contacts, that would be a snap. Instead, Rock wanted to find investors who could offer some expertise that would be useful to the new business. Partly out of a wish to prevent any single shareholder from becoming too powerful, he decided to scale back the contributions offered by his fifteen contacts. One of the new company’s founding shareholders was Grinnel College of Iowa, Robert Noyce’s Alma Mater. By now, the pig incident had long been forgotten. With Noyce now chairman of the college’s board of trustees, the success of Fairchild made it seem only natural to sell his old college a $300,000 slice of the new venture. The board was happy with the idea; its resident investment guru, Warren Buffett of Omaha, Nebraska, had no objection, and so the investment went ahead. Noyce and Moore had already each put up $250,000; Rock himself came in for $300,000. The backers’ investment in Noyce and Moore would be the best deal they ever made.
It took only a few weeks to choose a business name. The pair started out calling themselves ‘NM Electronics’, but it was agreed on all sides that this sounded kind of old-fashioned. After discarding more than a dozen other ideas, Moore finally came up with ‘Integrated Electronics’. Noyce admitted that it captured just what the new business was about, but suggested that they should contract the name to a single word: Intel. Arthur Rock, whose work in raising the company’s startup capital had earned him the job of Intel’s first chairman, had no objection. The pair discovered too late, shortly after the incorporation papers were drawn up on 16 July 1968, that there was already a company in existence called Intelco. By then, the word was already out about their new venture, and the easiest way to head off any lawsuits was to pay $15,000 for the right to use the Intel name.
Two weeks later, in August 1968, the local Palo Alto Times got wind of the story, and called up Gordon Moore for an interview. The resulting story, which ran across the top of an entire page of the paper, helpfully included both men’s home addresses. Within days, they began to receive résumés, calls and letters of application from talented engineers all over the industry who knew of Noyce’s work on the integrated circuit or of Moore’s achievements at Fairchild. Finding good employees willing to join a startup – usually one of the hardest tasks facing most entrepreneurs – was clearly not going to be a problem.
2
2,000 Starts a Week
GREAT CHANGES WERE HAPPENING in the world during the long, hot summer of 1968. In Paris, students set up barricades against their college professors and the government. In Chicago, demonstrations against the war in Vietnam turned into riots that marred the Democratic National Convention. But the foundation of Intel, which has had a bigger effect on the lives of people in the world than either of these, took place without being noticed outside the narrow confines of the Bay Area’s electronics industry.
That was just how Gordon Moore liked it. The very opposite of a self-publicist, he instinctively avoided talking about his achievements until they were complete.
When Moore was asked by one of the local papers at the beginning of August what kind of devices he and Robert Noyce intended to build at their new company, his response was about as uninformative as it could be: ‘We particularly are interested in the product areas that none of the manufacturers are supplying,’ he said.
Come on, Gordon, that’s hardly news. You wouldn’t be setting up a new company just to copy something already on the market, would you?
When asked to elaborate, Moore would offer only two further thoughts. First, the new company would avoid direct government business. Second, the company would keep to the industrial end of the business rather than trying to develop products for end consumers.
From anyone else, this apparent vagueness might have indicated indecisiveness. But Moore hadn’t spent nine years at the helm of Fairchild’s extensive R&D operation for nothing. His principle was straightforward: there was no need to give away any information, however trivial, to potential competitors. If the people who had backed him and Noyce to the tune of several hundred thousand dollars each could do without detailed information on their technical plans, then so could the rest of the world.