Our final installment explores the legacy of three key advanced that changed the world in the 1960s: the integrated circuit, startup fever, and venture capital.
In parts 1 and 2 of this three-part series, we recounted the birth of the transistor, William Shockley Jr.’s history in Silicon Valley, how Fairchild Semiconductor came into being, the evolution of planar technology, the “family tree” of semiconductor startups that evolved from Fairchild (the :”Fairchildren”), including Intel, and the competition with Texas Instruments. In part 3, we explore the legacy of Silicon Valley as a result of the three key inventions that changed the world in the 1960s: the integrated circuit, startup fever and venture capital.
Melting Pot for the Fairchildren
Sheldon Roberts, Eugene Kleiner, and Jean Hoerni’s collective decision to leave and compete against Fairchild, just over three years after the company was founded, was the first of what would be many subsequent defections and spinouts, eventually known as “Fairchildren,” directly or indirectly creating dozens of corporations, including Intel and AMD. In doing so, Fairchild sowed the seeds of innovation across multiple companies in the region that would eventually become known as Silicon Valley.
Local watering holes, restaurants and other hot spots provided venues for Silicon Valley’s “work hard, play hard” ethos, where industry folk gathered after work to drink, gossip, brag, trade war stories, talk shop, exchange ideas, change jobs and develop new contacts. Key venues included the Wagon Wheel, Lion & Compass, and Ricky’s, along with the Peppermill and the Sunnyvale Hilton.
Stanford University, and particularly Fredrick Terman, also played a catalytic role, propelled by the engineering department chair’s vision for academia to develop a new relationship with the science and technology-based industries dependent on brain power as their greatest asset. Terman further recognized the need to develop local industry, not just by building a community of interest between faculty and industry but also by encouraging new enterprises — what today we would call startups — to cluster around the university. To that end, Stanford provided intellectual property and office space, often rent-free other than the local property taxes.
While it is unclear who came up with the moniker, “Silicon Valley,” Don Hoefler, a technology reporter for the industry publication Electronic News, is often credited with popularizing the name in a 1971 column about the region’s chip industry. Hoefler also promoted the area’s innovative qualities, and was one of the first writers to chronicle the Northern Californian technology industry as a community.
The Fairchild Legacy
Throughout the first half of the 1960s, Fairchild was the undisputed semiconductor leader, forging ahead across all industry segments, be it design, technology, production or sales. Early sales and marketing efforts were modest and military-oriented; that changed in 1961 when Robert Noyce and Tom Bay recruited a group of aggressive salesmen and marketing specialists, including Jerry Sanders III and Floyd Kvamme. The newcomers transformed Fairchild’s sales and marketing departments into one of the industry’s legends.
Among the pivotal moments was Fairchild’s entry into the consumer TV market. Attracted by potential high volumes, Sanders wanted to replace the tube (valve) CRT driver with a transistor, but the target price was U.S. $1.50. Transistors at that time were selling to the military for $150.00. In what can only be regarded as a massive leap of faith, Noyce’s instructions to Sanders were, “Go take the order, Jerry. We’ll figure out how to do it later. Maybe we’ll have to build it in Hong Kong and put it in plastic, but right now let’s just do it.”
In 1963, Fairchild hired Robert Widlar to design analog operational amplifiers using Fairchild’s digital IC process. Despite its unsuitability, Widlar, in partnership with process engineer Dave Talbert, succeeded in adapting the process to produce two revolutionary parts: the world’s first operational amplifiers, the µA702 in 1964 and µA709 in 1965. With these two parts, Fairchild now dominated both the analog and digital IC markets, first with its µLogic RTL family and then with its 930 series DTL. In April 1965, Gordon Moore famously publishing his article “Cramming More Components Onto Integrated Circuits” in Electronics. Later to be known as Moore’s Law, it was basically an extrapolation of four plots on a graph showing IC transistor density over time.
Fairchild’s digital technology lead was, however, being overtaken by Texas Instruments. Having fallen behind in RTL and DTL, Fairchild’s chief rival decided to copy Sylvania’s ultra-high performance (SHUL) transistor–transistor logic (TTL) circuit design, adapting it to its own process to counter the announcement of Fairchild’s third generation 9000 series TTL logic.
Headed up by Stewart Carroll, Texas Instruments set up a “design factory” that could churn out several new designs a week, mostly by guessing the W/L ratios, laying out the circuits, correcting them if the prototypes did not work, and zeroing in on a specification that manufacturing could support. The design factory was supported by an optical photomask generator, as opposed to a manual rubylith layout, that could quickly create a photographic chip layout, and a “quick-turn” fab line to rapidly churn out parts.
TTL Data Book
To strengthen their attack, Texas Instruments masterminded a marketing coup by persuading other semiconductor firms to second source its TTL rather than Fairchild’s competing product. In this single masterly move, Texas Instruments established its 74 Series version of TTL as the de facto third-generation industry standard, leaving Sylvania’s SHUL, Fairchild’s 9000 series and other proprietary alternatives behind. It then proceeded to neutralize the entire second-source movement by providing every engineer with a copy of its ubiquitous orange book (The TTL Data Book), its twice-yearly “must-attend” TTL seminars, not just in the U.S. but globally, supported by an aggressive new product introduction program.
By always ensuring any bill of materials included at least one TTL part that was only available from it, Texas Instruments was able to stay one step ahead of the competition and own the TTL market for the best part of 30 years, until standard logic eventually fell victim to the 1980s application-specific IC revolution.
In the meantime, starved of capex, Noyce’s position on Fairchild’s executive staff was consistently being undermined by Sherman Fairchild’s corporate interference and his lack of support. The Fairchild management team was increasingly upset by Sherman’s corporate focus on unprofitable ventures at the expense of the semiconductor division. The company suffered its ultimate humiliation in July 1967 when the semiconductor industry fell victim to the first of its cyclical recessions, during which the company lost money and was forced to concede its technology leadership to Texas Instruments.
Charles Sporck, Noyce’s operations manager often credited with running the industry’s tightest ship, left in early 1968 along with Pierre Lamond to join Widlar and Talbert at National Semiconductor. That triggered Noyce and Moore’s departure from the firm later that same year–a pivotal moment in the eventual demise of the firm. The collective exodus of Sporck, Noyce, and Moore, along with so many other executives, signaled the end of an era, prompting Sherman Fairchild to bring in a new management team, led by C. Lester Hogan, then vice president of Motorola Semiconductor.
Of the eight original founders, only Julius Blank remained, although he too would be gone within a year.
Hogan’s arrival, and the subsequent displacement of Fairchild managers, demoralized the firm even further, prompting a further exodus of employees who would launch a host of new companies. Leading a group dubbed “Hogan’s Heroes,” the ultra-conservative Motorola executives immediately clashed with Sanders, Fairchild’s flamboyant sales chief.
While initially slow to respond to the changing market under Sander’s direction, Fairchild embarked on a strategy of leapfrogging Texas Instruments by focusing on more complex large scale, 30-plus gate parts, instead of simpler small and medium scale devices under 30 gates — a strategy that was proving popular and successful with engineers. The move forced Texas Instruments to recognize the threat and copy all of Fairchild’s 9300 series parts under 74 series numbers (for example the 9300 became the 74195 and the 9341 the 74181.)
Sander’s entire strategy collapsed, however, when Hogan capitulated to Ken Olsen, founder and CEO of Digital Equipment Corporation and a key Fairchild customer. Olsen wanted Fairchild to give up on its proprietary TTL technology and instead second-source Texas Instruments’ 74 Series TTL. Against Sanders’ wishes, Hogan agreed, signing the death warrant for Fairchild’s TTL strategy. Sanders was, understandably livid. “You’ve just killed the company, Ken,” Sander’s fumed.
Hogan’s betrayal was the last straw for Sanders. He, together with a group of Fairchild engineers, quit to start Advanced Micro Devices. With Sanders installed as president, one of his first moves was to establish the mantra: “People first, revenues and profits will follow.” Sanders also gave every employee stock options in the new company, an innovation at the time.
Wilf Corrigan, who had moved with Lester Hogan as director of Discrete Product Groups, succeeded Hogan as president and CEO in 1974. Fairchild continued to decline, however, dropping to sixth place in the semiconductor industry by the end of the decade.
In the summer of 1979, with the semiconductor market again riding high on its fourth year of successive double-digit growth, Fairchild fell victim to a hostile takeover bid from Gould, a major U.S. producer of electrical and electronic equipment, hell-bent on a diversification strategy.
Unable to fend off the buyout, Corrigan sought the best price for shareholders. Fairchild was eventually sold to Schlumberger, a French oil services industry company, for $350 million or $66 per share (Gould went as high as $57 per share).
Schlumberger was unable to revive the deteriorating company, and it continued to lose money. Corrigan departed in February 1980 and, once his non-compete clause expired, he and Rob Walker co-founded ASIC pioneer LSI Logic Corp. in 1981.
Schlumberger initially replaced Corrigan at Fairchild with one of its own managers, Tom Roberts, who unsuccessfully ran the firm like a heavy equipment company. Two years later, in 1983, it recruited Donald W. Brooks, a Texas Instruments veteran, to reverse its decline. By then, Fairchild was a legend in trouble, lagging in leading-edge technologies and losing money, even as the rest of the semiconductor industry was booming.
The company was eventually sold to National Semiconductor in 1987 for one-third of the price paid by Schlumberger eight years earlier. With the Fairchild brand now dead, Brooks left, and the company was back in the hands of former Fairchild General Manager Charlie Sporck.
Kirk Pond became COO at National Semiconductor in 1994, where he led the successful management buyout in 1997. With the Fairchild name revived, Pond continued as president and CEO until 2005, when he became chairman, before retiring a year later in 2006.
Pond was succeeded by Mark Thompson until the company was acquired by ON Semiconductor in September 2016. ON Semiconductor was the discrete, standard analog and logic device division spun out from Motorola’s Semiconductor Components Group in 1999.
Silicon Valley Legacy
Three critical advances in the 1960s–integrated circuits, startup fever and venture capital–changed the world. Perhaps these inventions would have happened somewhere else, at some other time, by somebody else. The fact they all occurred within a short space of time — in and around Palo Alto — driven by the entrepreneurial spirit of the “Traitorous Eight” and many other key players, along with the startup ethos of Stanford University, made Silicon Valley unique in the history of technology.
But what if Shockley’s parents had moved to Colorado, Nevada, or West Virginia to pursue their mining careers on their return to the U.S. rather than Palo Alto? Would Silicon Valley had developed instead in those places?
What if Shockley had chosen to set up Shockley Semiconductors on the East Coast, where there was an already well-developed infrastructure, rather than Palo Alto, which had none? From the perspective of infrastructure, the East Coast was far better positioned to forge a Silicon Valley.
What if the Russians, Europeans or Japanese had invented the integrated circuit first? Could Silicon Valley have sprung up in the USSR, Europe or Tokyo instead?
What if Fredrick Terman of Stanford University had not had the foresight to develop a community of interest between Stanford’s faculty and a fledgling industry, encouraging new enterprises to cluster around the university?
What would the world look like today had any of these scenarios not unfolded?
Clearly, fate played a role in bringing Shockley and semiconductors to Palo Alto, but the West Coast proved a far more fertile environment for the risk-taking spirit of the Traitorous Eight and their peers than the more risk-averse East Coast business and financial communities.
All eight of the original founders eventually left Fairchild and went on to become serial entrepreneurs, co-founding between them a wide variety of new startups, both in semiconductors and venture capital. They surrounded themselves with brilliant engineers who also wanted to start new companies, prove themselves, and change the world — stoking the startup boom driven by Shockley Semiconductor as the embryo, Fairchild Semiconductor as the incubator and the Palo Alto ecosystem as the catalyst.
The rest, as they say, is history.
On February 14, 1956, Arnold O. Beckman and William B. Shockley announced to a luncheon audience of scientists, educators, and the press at San Francisco’s St. Francis Hotel that they were founding Shockley Semiconductor Laboratory in Palo Alto.
The entrepreneurial spirit of the Valley, and the rise and fall of Fairchild, is best summed up by the following comment from Rob Walker, co-founder of LSI Logic: “It’s amazing what a few dedicated people can accomplish when they have clear goals and a minimum of corporate bullshit.”
This article was originally published on EE Times.
Malcolm Penn is the chairman, CEO, and founder of semiconductor industry analyst Future Horizons.