The Computer Chip is Sixty
How a signature Silicon Valley start-up’s success decimated the company but led to “the greatest legal accumulation of wealth in history”
Sixty years ago, start-up Fairchild Semiconductor of Mountain View, CA, introduced the first practical integrated electronic microcircuit to a gathering of press and industry insiders at the 1961 IRE (Institute of Radio Engineers) Convention in New York. Popularly known today as a computer chip, Fairchild’s announcement echoed far beyond the plush ballroom of the St Moritz Hotel (today the Ritz Carlton) overlooking Central Park.
Under the title “In Tiny Devices a Revolution,” a photo in LIFE magazine compared Fairchild’s new chip to the size of the letter “D” on a dime. It said “their job is to switch, amplify and alter electric currents just as transistors and vacuum tubes do, but one of them equals a dozen tubes plus yards of wiring.” The device contained just four transistors; today’s chips hold billions.
One of a handful of Silicon Valley innovations that genuinely “changed the world,” this tiny chip that crammed a complete electronic circuit onto a sliver of silicon also begat Fairchild’s demise. Internal battles over its future kicked off a cycle of spinouts, successes, and failures that destroyed the company but built Silicon Valley.
A “Tyranny of Numbers”
Following their introduction in 1948, computer designers quickly adopted transistors as replacements for bulky, power-hungry vacuum tubes. As systems grew in complexity, they faced a “tyranny of numbers” problem. Figuring how to connect hundreds of thousands of tiny transistor packages together in a computer became exponentially more difficult with each new generation of machines.
Between 1953 and 1955, scientists at research labs of prominent East Coast electronics companies, including Bell Telephone Laboratories, IBM, and RCA, built experimental devices that interconnected several transistors on a single semiconductor chip but offered limited functionality. Basking in the success of the first commercial silicon transistor, Texas Instruments (TI) of Dallas, TX, hired Jack Kilby to develop a more general-purpose solution. In 1958, he demonstrated, what he called, a “solid circuit.” Although it required careful handcrafting of microscopic gold wires, TI management saw the future and encouraged Kilby’s work.
Planar technology paves the way
Founded in 1957, Fairchild Semiconductor of Mountain View, CA, enjoyed instant success with an improved silicon transistor. Orders from US government contractors racing to overtake Soviet superiority in space with smaller lighter rockets gave the start-up immediate revenue and credibility. Reliability problems with the chips threatened to derail the company. Founder Jean Hoerni developed a solution called the “planar” process that coated the silicon surface with a thin glass layer (silicon dioxide). Licensed by every major semiconductor manufacturer of the era, planar remains the fundamental approach to making computer chips today.
Hoerni’s approach yielded other benefits. It allowed transistor patterns to be “printed” on a wafer using photolithographic masking layers — similar to screening colors on a T-shirt. This reduced transistor manufacturing cost as automatic machines replaced manual operations. Fairchild Director of R & D, Robert Noyce, also realized that metal traces deposited on top of the insulating glass layer could be used to connect multiple transistors in the underlying silicon wafer into a complete integrated electronic circuit (IC).
Aware of TI’s work, Noyce saw planar as a practical way to implement Kilby’s concept. Gordon Moore, who later succeeded Noyce as director, calculated that an IC could be cheaper to make than the sum of the individual transistors it replaced. Concerned that TI might beat Fairchild to the market, in late 1959, Noyce charged co-founder Jay Last with the task of developing an IC.
Making Micrologic
Last assembled a team to figure out how to translate Noyce’s idea into a manufacturable product. Overcoming significant technical challenges, equipment limitations, and naysayers within the organization, they raced to produce a commercial product they called Micrologic for the 1961 IRE show.
Aerospace system designers, including a group working on the Apollo Guidance Computer for the 1969 moon-landing, chose Micrologic devices to resolve size and weight constraints and placed orders for immediate delivery. As with many state-of-the-art IC designs, the low yield of early Micrologic chips consumed a significant portion of Fairchild’s manufacturing capability. At the same time, demand for popular and highly-profitable transistors stirred conflict over how to allocate scarce production capacity. Tom Bay, the influential head of sales, demanded that Last stop working on ICs to address improving transistor output. Last believed in his work’s future and lacking support from senior management he quit to form Amelco in 1961. Jean Hoerni and two other Fairchild founders joined him. A couple of months later, key members of the Micrologic team left to found Signetics to focus on ICs.
The exodus accelerates
As venture capital funds entered the arena, other Fairchild employees saw the opportunity to create their own companies and share the profits from a rapidly growing market. According to Moore, “It seemed like every time we had a new product idea, we had several spin-offs.” This debilitating talent drain combined with lack of investment by parent Fairchild Camera & Instrument Corporation weakened the Semiconductor Division.
Moore and Noyce delivered the most severe blow in 1968. They left to found Intel and in 1971 oversee the introduction of the microprocessor. Fairchild continued to grow but lost market leadership to TI. One of its progenies, National Semiconductor, acquired the company in 1987. In turn TI purchased National in 2011.
The drain continued with Advanced Micro Devices (AMD) and Monolithic Memories in 1969, followed by a host of second-generation spin-outs a decade later, including VLSI Technology (1979), Linear Technology (1981), LSI Logic (1981), Cypress (1982), Maxim (1983), and Altera (1983) who made important contributions to the development of new technologies and markets.
“The greatest legal accumulation of wealth in history”
In 2017, industry analyst Rett Morris of Endeavor Insight hailed Fairchild Semiconductor as “The First Trillion-Dollar Startup.” Through its technical, business, and cultural innovations, the company spawned hundreds of ventures (more than 126 semiconductor companies alone through 1986) that established Silicon Valley as a world center of entrepreneurial activity and technological leadership. Morris traced the lineage of Apple, Facebook, Google, and 89 other public Silicon Valley companies back to Fairchild. Although the firm’s market valuation never exceeded $2.5 billion, he estimated the surviving combined progeny’s worth at over $2 trillion.
Venture capitalist John Doerr has attributed this legacy of the unexpected fallout from the New York City launch of Micrologic sixty years ago this month as leading to the “the greatest legal accumulation of wealth in history.”
[Rev: 8.14.21]
