In October 1969, Business Week, as part of a special report on semiconductors, commissioned a cartoon that captured the spirit of the times. “The semiconductor industry is in a greater frenzy than ever before as companies vie for a slice of the lucrative market,” read the caption. At the far left of the drawing stands the headquarters of “Integrated Circuits, Inc.,” a low-slung building with a madly whirling revolving door, dozens of employees bumping into each other on every floor, and still other employees tossing a computer out a window. People who can’t fit in or out the revolving door (it’s quite full) are climbing into the building through windows, or else using the windows to jump to the ground and run to other integrated circuit companies. A man is being plucked from a wall by an automatic arm attached to the shell of a yet-to-be-built building that has a sign plastered to its side: “Roll Your Own Integrated Circuits.” There’s a shop called “Frontier IC,” a lean-to marked “Overnite Integrated Circuits Co.,” and an operation the size of a hot dog stand that calls itself “Instant IC Co.”
Meanwhile, in the far right corner of the drawing, salesmen are at battle. “Want a beam-leaded MSI Flip-Flop Chip in MOS-Compatible DLT off-the-shelf for $2.14 the unit at 10,000K?” one asks. “That went out back in July!” counters another. “We’ll give you …” Business-suited people run back and forth; a helicopter emblazoned “Sky Hook Integrated Circuits, Inc.” pulls employees from rooftops and chop-chop-chops them through the sky; heavy technical equipment is hauled up the sides of buildings.
The cartoon people look a little addled but perfectly happy. The comic depicts chaos to its readers, but for the folks in the picture, the scene is darn near heaven.14
Startup fever ran so high that the major industry trade show in 1969 offered several sessions on management problems, starting companies, and talking with venture capitalists. Attendance at a session called “New Company Start-ups: The Engineer Becomes Entrepreneur” broke all previous records, with nearly 1,000 people either squeezed into the conference room (many sat on the floor or stood in the back) or watching the closed-circuit television broadcast on a stifling August afternoon.15
Just as the industry’s high hopes for integrated circuits had launched the earlier gaggle of startup companies, the 1968–1969 generation was inspired by a belief that semiconductors were on the cusp of another dramatic technological breakthrough. In 1965, Gordon Moore had pointed out that every 18 months, the number of transistors on an integrated circuit doubled. If this trend predicted by “Moore’s Law” continued, the day was not far off when integrated circuits could be orders of magnitude more complex than their counterparts just a few years ago. Already Moore’s own R&D group at Fairchild had fit a once-unthinkable 1,024 transistors onto a single circuit. In 1968, this circuit was little more than a lab curiosity, but the general consensus held that circuits with more than 1,000 components integrated together—so-called Large Scale Integrated circuits—should be physically possible to mass produce by 1970.16
This era of Large Scale Integration (LSI) would usher in integrated circuits that could not only perform their current functions more quickly and cheaply than ever before but could also do things that had previously been considered too complex or expensive for semiconductors, such as serve as memories for computers or as control drivers for heavy machinery. LSI also promised a boost for a semiconductor company’s bottom line because a single chip integrating thousands of functions could drastically cut the cost-per-function while still commanding top prices in the market. Noyce predicted that the economic impact of LSI would eventually rival that of the integrated circuit itself.17
Most of the new semiconductor companies formed at the end of the 1960s, including Noyce and Moore’s startup, were begun in hopes of bringing LSI circuits to market before anyone else. Noyce and Moore, in fact, had settled on computer memories as a first product not primarily because the computer market was growing—although that was a welcome reality—but because memories would be the easiest types of LSI circuits to build. Memories consist of row upon row of identical transistors, laid out in a gridlike pattern that was far easier to design and manufacture than the mazes of gates and leads used in logic circuits. Moreover, the regular layout pattern cut down on the interconnection and packaging costs that often represented the most expensive and complicated parts of a circuit. In Moore’s words, LSI was a “technology looking for applications” in 1968. Which is to say: if LSI technology was going to work anywhere, it would work first in memories. If it worked in memories, Noyce and Moore could anticipate an ever-growing market of computer makers ready to buy.18
“I KEEP REALIZING that today your office is at home for the present,” wrote Noyce’s mother Harriet in July of 1968. She no doubt imagined it quite a come-down from an executive suite at Fairchild Camera and Instrument. “What a lot of thinking about the future and dreaming and playing with a lot of ideas you must have done already. It is a courageous step.” In her inimitable fashion she added, “Well, we have no conception how you organize a start. I hope that it is exciting to you and in some way exhilarating, and that you don’t feel pushed or hurried into organizing a company before you are ready.”19
Noyce was more than ready. He threw himself into the launch of this new company with a professional enthusiasm he had not felt for years. He talked to bankers, attorneys, journalists, realtors, equipment suppliers, possible board members, architects, insurance agents, and potential customers. He met several times with his personal accountant. He was happy handling almost any aspect of the company, from drawing up a business plan (“planning, cash flow, how big a hole[?] Cap Equip, People, Revenue, Guess of balance sheet”), to re-immersing himself in technical discussions, to meeting with a fellow who organized softball teams for a popular corporate league, to typing documents and securing tax identification numbers for the new venture.20
People from Fairchild were continually resurfacing at this stage, and not just among the inner circle in Noyce’s study. When Noyce and Moore found a building they thought might work to house offices and a fab, Julius Blank, still at Fairchild, dropped by to help them evaluate the site. Noyce’s contact at the semiconductor company vacating the facility was Dave Beadling, who had once headed Fairchild’s work on the Minuteman missile. The company for which Beadling now worked, the semiconductor division of Union Carbide, had been started by Jean Hoerni after he left Jay Last and Amelco-Teledyne. In July and August alone, Noyce noted meetings with a dozen current or former Fairchild employees, most of them looking for jobs. Chances are high that meetings with many other Fairchild people went unrecorded.
The West Coast semiconductor industry was still so small that it seemed as if nearly everyone in a professional position knew everyone else—often from having worked together at Fairchild. Jay Last’s Amelco-Teledyne rented space from Ed Baldwin’s Rheem, a company later bought by Raytheon, yet another Fairchild spin-off, which years later would hook up with former Fairchild CEO John Carter, who had started his own new company. The industry watering holes—Rudy’s, the Wagon Wheel, Dinah’s, Chez Yvonne, the Velvet Turtle—were as busy as ever, and the fact that two guys no longer worked at the same company did not keep them from grabbing a table and discussing subjects that would have given corporate attorneys fits. Raiding became a high art, for as Noyce put it, “the schools weren’t turning out anyone that knew anything about [semiconductors, and] consequently the only source of knowledgeable people were the companies that were already working in the field.” It seemed people had a greater sense of loyalty to the industry itself than to any particular company.21
A key part of Noyce’s early work at NM Electronics involved convincing the most capable engineers, technicians, and scientists he knew to join the company. “We are only going to hire perfect people,” Noyce told his oldest children one summer afternoon. “A small bunch of people who know what they are doing can accomplish much more than a big group of people who don’t know what they are doing.” When Noyce attended the Fall Joint Com
puter Conference in San Francisco he was nearly overrun by men wanting jobs and asking questions. At one point during the conference, he apparently had meetings scheduled with so many Fairchild employees that he simply noted in his datebook, “Fairchild Circus.”22
Within a month of launching NM Electronics, Noyce was on the East Coast, recruiting for his new venture. “He was like the Pied Piper,” recalled Roger Borovoy, attorney at both Fairchild and Intel. “If Bob wants you to come, you come.” Early recruiting advertisements requested that applicants “please drop a note with qualifications to Bob Noyce” adding oh-so-casually, “he is still doing our personnel work.”23
Noyce also hunted for new talent at Stanford. Jim Angell, a friend from MIT and Philco who now taught in the university’s engineering department, invited Noyce to address his students. Noyce also told Angell that he was looking for “a good circuits guy who also knows [computer] systems.” Noyce needed someone familiar with computers to help determine the most desirable features of semiconductor memories. Angell, himself a systems expert, suggested that he join the company, but Noyce refused to consider pulling him from academics. He had a great respect for teachers and thought Angell was one of the best he had ever seen. And so Angell suggested that Noyce talk to one of the postdoctoral fellows in the department, a young man so gifted with computers that Angell swore he could tell whether a program was running properly by the rhythm of the lights on the display.24
Ted Hoff, who would soon be known around the world as the inventor of the microprocessor, came for a job interview at Noyce’s study shortly after Noyce and Moore incorporated their company. Hoff had heard rumors about Noyce, the most persistent of which claimed that Fairchild had made him “a millionaire or close to it.” One look at Noyce’s house was enough to convince Hoff of the rumor’s veracity. “If you’re in academia and you do something good, you get a nice pat on the back,” Hoff thought, as he made his way to Noyce’s front door. “When you’re in industry and you do something good, people throw money.”25
Noyce led Hoff to the study and then asked him the central question of the interview: what did he think would be the next big area for semiconductors? Hoff immediately answered “memories,” though he had no idea that this was the area Noyce and Moore sought to target. Hoff then asked Noyce if the world really needed another semiconductor company. Several of the Fairchildren had failed rather spectacularly, and Hoff wanted to know why this venture would not follow suit.26
Noyce had asked himself the same question. He had briefly wondered if he and Moore were too old—Moore at 39, Noyce at 40—to start a company. Moreover, as much as Noyce longed to “get close to advanced technology again,” he was concerned that perhaps he had “been away [from the lab] too long” to jump back into the game. After a bit of thought, however, he had decided that his and Moore’s age, if reconsidered as experience, was an asset. As he put it, “the semiconductor business hadn’t existed longer than we’d been in it. There wasn’t anybody who knew the business better than we did.” Noyce was also confident that Moore’s familiarity with state-of-the-art advances in the technology more than compensated for his own relative lack of knowledge. And although he was almost loath to admit it, the contacts and skills Noyce had acquired in his managerial work at Fairchild had their own value. Besides, if for some reason this new firm faltered, Noyce figured he and Moore could sell it to a computer company.27
In fact, Noyce told Hoff at his job interview, even if the company was not a runaway hit, the founders and early employees could expect to “do quite well,” thanks to the stock options they would all hold. After consulting an attorney to find out what, exactly, a stock option was and whether it was potentially valuable, Hoff decided to join Noyce’s and Moore’s team in the nebulously defined position of “manager of applications research.” His reasoning? “I felt I was young enough that this wouldn’t be my last job opportunity. If this panned out, great. If not, there would be other opportunities.” Hoff may have been too young to know it at the time, but this cavalier attitude was a direct outgrowth of Noyce, Moore, and their six co-founders’ decision to depart Shockley a little more than a decade before. In 1957, Arnold Beckman could seriously call leaving an established company for a more attractive alternative “disloyal” and “shameful.” Now it was routine.28
By the time Hoff and Noyce talked, summer vacation had begun, and the Noyce children, who had not yet left with their mother for their annual summer pilgrimage to the East Coast, were everywhere—in the house, in the yard, swinging on the rope over the pond, or chasing after the pony, who forever seemed to be nudging her gate open and cantering down the street. Added to these distractions was a near-constant stream of visitors and ceaseless ringing of the telephone. When it got to be too much for Noyce and Moore, they would wander through their neighborhood—they lived close to each other in Los Altos—or even commandeer a neighbor’s porch in hopes of finding a quiet place to talk and think.29
Noyce and Moore spent a good deal of time considering a name for the company. Noyce wanted a moniker that was “sort of sexy,” a criteria not met by NM Electronics, nor by Moore-Noyce, an early contender that suffered from the problem of sounding like “more noise”—not an ideal association for an electronics operation. Noyce and Moore considered various combinations of California, Electronic, Computer, and Technology. When none of these choices were available—at least four had recently been taken by other new companies—Noyce and Moore drafted a list of some 20 other possibilities, including “Electronic Solid [S]tate Computer Technology” (abbreviated to “Esscotek”), “Electronic Computer” (abbreviated to “Tronicom”), and “Integrated Electronics” (abbreviated to “Intel”).30
Noyce liked “Intel,” which he felt met the “sort of sexy” criteria because it “implied other things, rather than just another company”—precisely what other things remains unclear, although the association with “intelligence” was obvious. The attorneys found a few other “Intels,” including a hotel supply chain in Ohio and an International Television Company in New York, but there were no conflicts that could not be resolved by buying out a license. The new company would be called Intel.31
IF NOYCE AND MOORE had a third co-founder in the launch of Intel, it was financier Arthur Rock, who had been urging Noyce to leave Fairchild long before Noyce actually decided to do it. When he began working on the Intel launch, Rock’s partnership with Tom Davis had just dissolved, under the terms of its own contract and with great success. Investments made with the first $3 million of the fund (the other $2 million were never invested) were now, seven years later, worth nearly $100 million, thanks in large measure to stakes in Teledyne and Scientific Data Systems, an early scientific computer company. Over the years, Rock had focused on investments in technology, but he considered his real expertise to be identifying good teams with the ability to change their industries and the desire to generate large profits. “The man has to have a killer instinct,” Rock said. “He has to know where the game is—on the bottom line.” Noyce personally thought that Rock’s own competitive intensity accounted for much of the success of the companies he funded. “The main thing is,” Noyce explained, “Art likes to win.” Noyce, who also hated to lose, admired and worked well with Rock.32
The two men spoke daily in the first weeks and months of Intel’s existence. Rock had decided to fund the company by selling 500,000 shares in it at $5 apiece. The formal investment vehicle would be $2.5 million of convertible debentures. Convertible debentures, which Rock had used in other funding situations, are a sophisticated IOU. Investors are treated like creditors and are paid interest on their loan—though Intel waived such payments—until the investors convert the debt into equity (stock). If the company went under, the early investors were at the end of the line of creditors. If, on the other hand, Intel proved successful, the first-round investors would together own half the company.33
Noyce and Moore wanted any potential investor to know that the company intended
to fund both a stock option plan and a stock purchase plan for employees. The founders believed that stock ownership was the best guarantee of both loyalty and innovation. Noyce thought profit sharing encouraged employees to stick with safe products already proven to generate a profit, however small. Owning stock, on the other hand, gave employees an incentive to pursue high-risk, high-reward, next-generation products—precisely the attitude Intel hoped to foster.34
Noyce had celebrated his freedom from Sherman Fairchild’s muttered curses about “creeping socialism” by beginning to outline stock-option and stock-purchase plans for Intel within days of leaving Fairchild. Rock also strongly supported a generous distribution of stock options. Scientific Data Systems, the biggest hit of his career thus far, had given options to almost all employees.35
The dissolution of Fairchild convinced Noyce, Moore, and Rock to design a stock-option plan that discouraged employee defections. “There are too many millionaires who did nothing for their company except leave after a short period of time,” Rock noted before proposing that Intel reserve for itself the “option to purchase stock of an employee who quits the company or is fired within a short period of time.” Rock also suggested that Intel “wait four to eight weeks before granting any stock options,” just to be safe in the churning market for semiconductor expertise. Such practices, which became industry standards, were important additions to the standard incremental vesting procedure already in use at Fairchild and other companies.36
Intel set aside options for 100,000 shares to be granted to key employees at a price of $5 per share. On the first day of 1969, a stock-purchase plan went into effect allowing full-time exempt employees to take up to 10 percent of their pay in Intel stock, at the same $5-per-share price through payroll deductions. For at least the first year of Intel’s existence, every eligible employee in the company elected to participate in the stock-purchase plan. By the end of 1968, options on 64,700 shares had been granted, and Intel had reserved an additional 25,000 shares for employee purchase.37
The Man Behind the Microchip Page 25