Chapter 1The Semiconductor Man

Foreword

In the summer of 1997 a man packed up an office in Cupertino, having run one of the most famous companies in the world for precisely five hundred days, and drove home to a life that no longer had an executive floor in it — and the industry he left behind closed over the space he had occupied so smoothly, so completely, that within a few years its official histories would treat those five hundred days as an interval rather than an administration, a dark stretch between two lit ones, the gap year the story politely skips.

That phrase — the gap year — is why I wanted to write about Gil Amelio, and it is also, I should admit at the outset, entirely outside the territory I know. My subjects have always been the British and American midcentury: the film studios and the repertory theaters and the television companies, and above all the people who ran those institutions in the intervals nobody commemorates. The artistic director who held a company together between two celebrated ones. The studio head remembered, if at all, as a trivia answer. The controller who commissioned the series that made his successor's name and was gone before it aired. I have found, over the years, that the interval is where the actual work of an institution tends to get done, and that the program note is the last place on earth you will ever read about it.

Amelio belongs to that company of caretakers, even though he never set foot in a theater I have written about and spent his career in an industry — Silicon Valley semiconductors and personal computers — that I came to knowing next to nothing. What drew me was not the technology. It was the shape of the man, which I recognized before I knew a single particular: the competent professional whose competence is the precise thing the myth around him requires to be erased, because a legend of the lone visionary cannot afford a predecessor who was merely good at his job.

I expected, when I started, to find a caretaker in the smallest sense — the man who kept the lights on, signed the checks, and had the good fortune to hand the keys to a genius. The record does not support that story. I worked from what anyone can work from: Amelio's own 1998 memoir, written in the flat, defensive, scrupulously fair prose of a physicist who believed he had been misread; the proxy statements and the financial disclosures; the trade press of the period, which was hungry for a hero and had already chosen one; and the slower, quieter reassessments that the industry's own historians have produced in the decades since. Read together, and against the grain of the mythology most of them were written to serve, they describe not a caretaker at all but something closer to a surgeon — a man brought in to stop a company bleeding to death, who did the unglamorous, unrepeatable work of stabilization, and who made, almost in passing, the one strategic decision that everything the company later became was built upon.

That decision, and the five hundred days around it, are the subject of this book. I have tried throughout to hold two things in my hands at once, which is the only honest way to write about a caretaker: that the man who comes after may genuinely be the greater artist, and that the institution he inherits was kept alive, and pointed in the right direction, by someone the greater artist's admirers would rather forget. Both of those things can be true. In Amelio's case I have come to believe they are.

It is an unfashionable kind of book to write, about an unfashionable kind of figure — the professional rather than the prophet, the interval rather than the première, the stabiliser in an industry that reserves all its language for transformation. But the interval is where I have always looked, and I have never once found one that was empty. This one, when I opened it, turned out to contain a foundation.

— Donovan Hadley, Manchester

Chapter 1 · The Semiconductor Man

Gil Amelio was not the kind of man who got written about. He was the kind of man who got things done, quietly, methodically, with the engineer's preference for function over form and the executive's understanding that the most important decisions are the decisions that nobody notices because they prevent the catastrophe that had otherwise made the evening news. His career was a long sequence of solved problems, of companies restored to health and industries navigated through transition, and the invisibility of that career was not accidental but structural: the culture in which he operated reserved its attention for the dramatic and the theatrical, for the product launch and the stock price surge and the visionary proclamation, and Amelio was constitutionally incapable of providing any of these things.

Gilbert Frank Amelio was born on March 1, 1943, in New York City, and the trajectory of his early life was the trajectory of postwar American technical meritocracy: public schools, a bachelor's degree in physics from the Georgia Institute of Technology in 1965, then a master's degree in 1967 and finally a doctorate in solid-state physics from the same institution in 1968. Georgia Tech was the kind of institution that shaped a particular type of American professional: rigorous, practical, oriented toward applied knowledge rather than theoretical speculation, the school where you went if you wanted to build things rather than merely contemplate them, and its graduate program sharpened those tendencies into research precision. By the time Amelio emerged with his doctorate, he was equipped with a powerful set of tools: the ability to understand complex physical systems at a level of mathematical rigor, to formulate hypotheses and test them against empirical data, and to derive practical applications from theoretical principles.

At Bell Labs, Amelio was part of the team that built the first working charge-coupled device, the CCD, in 1970, working alongside Michael F. Tompsett under the Bell Labs researcher George E. Smith. The device itself had been conceived the year before by Smith and Willard Boyle. The invention was, in its historical consequence, staggering. The CCD became the technology underlying every digital camera, every smartphone camera, every security camera, every barcode scanner, every astronomical telescope, every medical imaging device that relied on electronic light sensing. The invention of the CCD was not a minor contribution to a marginal problem. It was one of the foundational inventions of the digital age. George E. Smith and Willard Boyle would share half of the Nobel Prize in Physics in 2009 for the invention of the CCD. Amelio, who had contributed to its early development, received no Nobel. He received sixteen patents over his early career and became, in the accounting that Silicon Valley never made, one of the technologists whose inventions shaped the modern world. Instead, it became a footnote to a career that the Jobs mythology would reduce to a punchline.

He began his career at AT&T Bell Laboratories, and the entry point was consequential in ways that extend beyond the immediate opportunity. Bell Labs was the most productive research institution in the history of American technology: the place where the transistor was invented, where information theory was formulated, where the laser was developed, where the foundations of the digital age were laid not by individual geniuses but by a culture of systematic, collaborative, rigorous inquiry that assembled talent from every corner of American science and gave it resources, autonomy, and the institutional patience to pursue problems across decades rather than quarters. Amelio absorbed the Bell Labs culture, its commitment to doing things correctly rather than quickly, its understanding that the most important technological problems required sustained attention rather than inspired guessing, its deep respect for the difference between elegant solutions and expedient ones. He would carry that culture with him through everything that followed, and it would serve him well in every setting except the one in which Silicon Valley's theatrical expectations demanded that he perform the role of visionary genius.

From Bell Labs he moved into the semiconductor industry at Fairchild Semiconductor, and the choice of employer was, in retrospect, as consequential as any he made. Fairchild was the seedbed of Silicon Valley, the company from which the major semiconductor firms of the next generation sprouted like offshoots from a root system. Intel was founded by Fairchild alumni. AMD was founded by a Fairchild alumnus. National Semiconductor hired from Fairchild's ranks. The genealogy of Silicon Valley traces back to Fairchild the way the genealogy of Renaissance art traces back to Florence: it was the place where the skills were developed, the relationships were formed, and the culture was established that would define an industry. The culture of Fairchild Semiconductor was a culture of intense technical competition, of engineers who viewed every design problem as a personal challenge, of a meritocracy so pure that status was determined almost entirely by the quality of your circuit design and the ingenuity of your manufacturing process.

Amelio was, within this genealogy, a figure of substance but not of legend. He was not Robert Noyce, the charismatic co-inventor of the integrated circuit whose presence could fill a room and whose vision could redirect an industry. He was not Gordon Moore, whose eponymous law became the organizing principle of the semiconductor industry for fifty years. He was not Andy Grove, whose paranoid management philosophy ("only the paranoid survive") defined the culture of a company that would become the dominant semiconductor firm in the world. Amelio was an engineer who became a manager who became an executive, a man whose career ascended through the ranks of the semiconductor industry with the steady, unremarkable progress that characterized the careers of thousands of capable people in an era when the industry was creating wealth at a pace that the American economy had never seen.

The unremarkability is important. It is the background against which the Apple story must be understood, because the Apple mythology (the mythology of Steve Jobs, of visionary genius, of the lone creator who saw the future and willed it into existence) requires people like Amelio to be unremarkable. The mythology needs its foils to be mediocre, because if the foils are competent, then the hero's narrative becomes more complicated. If Amelio was a capable executive who made rational decisions under difficult conditions, then Jobs was not the sole savior of Apple but one participant in a complex institutional process. And that complexity is fatal to the mythology.

Amelio moved from Bell Labs to Fairchild to Rockwell International, accumulating experience in semiconductor manufacturing, research management, and corporate leadership. At Rockwell, he rose to the position of president of the communications systems division, managing a business with billions of dollars in revenue and thousands of employees. The position was, by the standards of corporate America, a significant achievement: the kind of role that required not just technical competence but the political skills, the financial acumen, and the managerial judgment that large-scale corporate leadership demands. Rockwell was not a glamorous company in the Silicon Valley sense; it was a defense contractor and industrial manufacturer whose semiconductor operations were one division among many, and managing that division required Amelio to develop the skills of a general manager (to understand supply chains and customer relationships and financial reporting and organizational behavior) alongside his technical background.

In February 1991, Amelio became president and CEO of National Semiconductor, taking over from the retiring CEO Charles Sporck. His name was already on sixteen patents, a record of technical contribution that distinguished him from the purely managerial executives who populated most corporate suites. The company was in perilous health. National Semiconductor had lost $320 million over the four years preceding Amelio's arrival, hemorrhaging money in a market where Japanese semiconductor companies (NEC, Toshiba, Hitachi, Fujitsu) had deployed manufacturing scale and government support to undercut American chip makers on price while matching them on quality. National Semi was a major player in the industry, but by the early 1990s it had fallen behind its competitors across virtually every dimension that mattered: its product lines were aging, its manufacturing processes were inefficient, its quality was inconsistent, and its financial performance was deteriorating toward the point where analysts were openly speculating about its survival as an independent company. Amelio's mandate was turnaround: cut costs, streamline operations, refocus the product portfolio, restore quality, and return the company to profitability.

He did it. The scale of the problem was staggering. National Semiconductor was producing over thirty-four thousand distinct products: an absurd sprawl of the product catalog that reflected decades of saying yes to customer requests without the discipline to ever say no, accumulating products the way a packrat accumulates objects, until the sheer weight of the collection crushed the organizational capacity to manage any of it well. When you manufacture thirty-four thousand products, you cannot manufacture any of them with the focus and the quality that each deserves. You spread engineering resources too thin. You fragment manufacturing capacity across too many configurations. You create a supply chain so complex that it cannot be managed efficiently. You produce nothing at the scale needed to achieve the cost curves that make commodity products profitable.

Amelio's diagnosis was precise and his prescription was radical: the company was trying to be everything to everyone, manufacturing commodity chips where it could never achieve the economies of scale that Japanese competitors possessed, and it was doing so at the expense of the areas where it actually had competitive advantage. His first year was what he would later describe as a product line massacre: he cut the thirty-four-thousand-product catalog to roughly fourteen thousand, killing every product line that could not demonstrate either market leadership or a clear path to thirty percent or better gross margins within eighteen months. Entire divisions were shuttered. Long-standing relationships with customers who wanted commodity products were sacrificed. Engineers who had spent years on products that were technically interesting but commercially marginal were redeployed or let go.

The analog bet was the strategic insight at the heart of the turnaround. While the rest of the semiconductor industry was obsessed with digital (with microprocessors and memory and logic chips, with the products that powered the computer revolution) Amelio recognized that analog chips, the components that translate real-world signals into digital data, occupied a different competitive environment. Analog chips required deep engineering expertise. They were harder to copy. They carried gross margins of fifty to sixty percent rather than the twenty to thirty percent typical of digital commodity chips. And the demand for them was growing as industrial automation, automotive electronics, and consumer devices increasingly required the ability to sense and measure the physical world. Amelio bet the company on analog, on the technical capabilities that National Semiconductor's engineers possessed and that could not be replicated overnight by competitors with large manufacturing plants and low labor costs.

The headcount reduction was the hardest part. Amelio laid off approximately six thousand employees between 1991 and 1993. He did not do this from a distance. He personally visited every affected site (the factories in Santa Clara, the design centers in Georgia and Texas, the sales offices around the country) and he met with the workers who were losing their jobs. He understood that each layoff was a person, a person with a mortgage, with children, with a career that had been built on the assumption that National Semiconductor was a company where talented people could do meaningful work indefinitely. The assumption was broken by the market forces that Amelio had not created, but the breaking fell on his account, and he knew it.

He formalized his turnaround methodology into what he called the Five Pillars framework, a conceptual structure that organized the interdependent elements of corporate recovery into five categories. The first pillar was financial restructuring: stop the bleeding, reduce the cash burn rate to a level that the company's revenue could sustain, create the financial breathing room that made everything else possible. The second was product rationalization: apply rigorous criteria to every product in the portfolio and eliminate everything that didn't meet them, concentrating engineering and manufacturing resources on the products that could achieve leadership positions. The third was cultural transformation: replace the defeatism and the learned helplessness that characterized an organization that had been in decline for years with a culture of accountability, of high expectations, of the belief that quality was achievable and that the company's future was in its own hands. The fourth was customer intimacy: rebuild the relationships with key customers that had been damaged by years of poor delivery, inconsistent quality, and the general unreliability of a company in decline. The fifth was technology investment: protect research and development even while cutting everywhere else, because the cutting was to preserve the future, not to destroy it, and destroying research capability was destroying the future.

He applied all five pillars simultaneously and with the discipline that only a physicist-turned-engineer could bring to an organizational intervention. By fiscal 1994, National Semiconductor posted net income of $264 million on revenue of $2.3 billion, a transformation from $152 million in losses that was, by any standard, extraordinary. The stock had risen from approximately five dollars per share when he took the company to over twenty dollars by 1994, and to above twenty-four when he departed for Apple in early 1996. The Fairchild Semiconductor division, which Amelio had ring-fenced from National Semi's higher-margin analog operations, laying the groundwork for its eventual independence, was spun off in March 1997 (just as his Apple tenure was ending) and unlocked roughly five hundred and fifty million dollars in shareholder value. Industry observers who had described National Semiconductor as "an unreliable third-tier chip producer" in 1991 were calling it "a respected front runner" five years later.

Apple was a major National Semiconductor customer. Amelio had joined Apple's board of directors on November 9, 1994, recruited by the headhunting firm Heidrick & Struggles, which had identified him as the kind of experienced technology executive whose manufacturing and supply chain expertise would be valuable to a company that was, among other problems, experiencing severe quality issues across its product line. His first board meeting featured a cryptic voicemail from Steve Jobs requesting a meeting, a call that went unreturned, an early signal of the collision that would define both men's legacies. Amelio was shocked by what he witnessed as a board member: the directors exercised virtually no control over CEO Michael Spindler, and at his second board meeting, Spindler stunned the room by announcing flatly that the company needed to be sold. October's fiscal year results brought catastrophe: a quarter that was supposed to produce $150 million in profit instead delivered a $69 million loss. Apple had over $1 billion in unsold merchandise sitting in distribution channels: a billion dollars of inventory that the market did not want at prices that did not cover manufacturing cost.

Peter Crisp, the Venrock founding partner who sat on the board, proposed Amelio as the next CEO in late December 1995. The search was brief. Amelio was the obvious candidate: he had turned around a comparable company, he was an engineer with credibility in Silicon Valley, he had been on Apple's board long enough to understand the dimensions of the crisis, and he was, critically, available. He negotiated his terms: a $200,000 raise over his National Semiconductor compensation, stock options, one million shares of Apple stock, and a $5 million loan. The loan would later be used against him by critics who characterized it as self-enrichment during a corporate crisis. In context, it was a standard element of executive compensation for a CEO taking an extraordinary risk. No rational person left a company he had successfully turned around to take command of a company that was, by multiple credible estimates, ninety days from insolvency, without compensation commensurate with the risk.

The board's logic for choosing Amelio was coherent given what the board believed. They believed Apple's disease was operational: too many products, too much overhead, a bloated cost structure that could be fixed by the same discipline Amelio had applied at National Semi. They believed that the turnaround playbook (cut products, cut costs, focus on margins, rebuild culture) would work at Apple as it had worked at National Semiconductor. They were right about the playbook. They were wrong about the disease.

What the board had failed to diagnose, what Amelio himself would not fully grasp until he was deep inside the company, was that Apple's crisis was not operational in the National Semiconductor sense. National Semi's problem had been a commodity company in a commodity market that had lost pricing power and was bleeding fixed costs against declining revenues. The prescription was focus and cost discipline. Apple's crisis was different in kind: it was a crisis of identity, of purpose, of the relationship between a company and its products and its customers. Apple had not merely become inefficient. It had become confused. It had lost the sense of why it existed, what it was for, what made a Macintosh worth more than a Windows PC, not technically (the technical argument was increasingly difficult to make) but aesthetically, culturally, experientially. The company had drifted from its founding conviction that technology should be beautiful, that computers should be designed for humans rather than for engineers, that the experience of using a computer should be a source of pleasure rather than mere utility. No amount of operational restructuring could restore that conviction. It required a visionary, not a surgeon.

Amelio was not a visionary. He was a surgeon. And what Apple needed, in February 1996, was first a surgeon and then a visionary. The surgeon had to come first. Because you cannot transform a patient who is dead.

The National Semiconductor turnaround was the credential that brought Amelio to Apple's attention as a CEO candidate. In 1996, when Apple's board of directors was looking for a CEO who could save a company that was widely described as ninety days from bankruptcy, the board was looking for a turnaround specialist: a person who had demonstrated the ability to take a failing company and make it functional. Amelio had demonstrated that ability. The demonstration was recent, credible, and relevant.

What the board was not looking for, what, in the conditions of February 1996, it could not afford to look for, was a visionary. Apple did not need a visionary. Apple needed a surgeon. The patient was bleeding out. The surgery had to happen immediately, and it had to be performed by someone who knew how to stop the bleeding, not by someone who had a vision of what the patient might become after the bleeding stopped.

This distinction (between the surgeon and the visionary, between the man who saves the patient and the man who transforms the patient, between the necessary and the spectacular) is the distinction that the Steve Jobs mythology erases. The mythology says that Apple was saved by Jobs. The financial record says that Apple was saved by the man who stopped the bleeding, and the man who stopped the bleeding was Gil Amelio.

The psychological dimension of the board recruitment process deserves attention because it illuminates something about Amelio's character that the mythology's caricature obscures. A man who had successfully turned around National Semiconductor (who had taken a company from $152 million in losses to $264 million in net income, who had watched the stock rise from $5 to $24, who had built a credibility and a reputation that placed him at the top of the technology industry's turnaround specialists) did not need to take the Apple assignment. He could have continued at National Semiconductor, could have accepted any of the other CEO positions that were being offered to successful technology executives of his stature. He had options. He chose Apple.

He chose it because the problem was interesting (a physicist's reason, the attraction of a genuinely difficult problem that would require the full deployment of every skill he possessed). He chose it because Apple mattered in a way that National Semiconductor, whatever its financial turnaround, did not. Apple had shaped the personal computer industry, had produced products that changed how people related to technology, had built a brand that carried genuine cultural significance beyond its market capitalization. He chose it because the people who approached him (Peter Crisp, the Venrock venture capitalist who had served on Apple's board and who believed Amelio was the right person for the job) were credible and persuasive. And he chose it, perhaps, because he understood that the failure to try was itself a kind of defeat, that the man who had helped build the first charge-coupled device and turned around National Semiconductor was not the kind of man who declined difficult problems because they were difficult.

Amelio was appointed CEO of Apple Computer on February 2, 1996. He was fifty-two years old. He had spent more than three decades in the semiconductor industry. He had a doctorate in solid-state physics and sixteen patents. He had turned around one of America's largest semiconductor companies. He had been on Apple's board for more than a year, long enough to understand that what awaited him was not merely difficult but potentially impossible.

He took the job anyway. The semiconductor man had arrived.

And the mythology machine had already begun to grind.