Mass Production Challenges and the Shift to Civilian Markets

Key Insight

Achieving mass production for integrated circuits required pioneering manufacturing techniques. Jay Lathrop at Texas Instruments developed 'photolithography' in 1958, a process using light and photoresists with an upside-down microscope to print miniature, precisely shaped transistor patterns onto semiconductor material, enabling features as small as 0.0005 inches. This innovation was vital for fulfilling the massive demand from military programs like Minuteman II and Apollo. Its implementation, however, necessitated TI to vertically integrate, producing its own ultra-pure photoresist chemicals, precise masks, and silicon wafers, as existing suppliers could not meet the required standards.

Mass production of chips was plagued by challenges, including chemical impurities, temperature and pressure variations, and dust contamination, all of which could ruin entire production runs. Improvement relied heavily on trial and error, involving thousands of experiments to assess various parameters. Engineers like Mary Anne Potter ran round-the-clock tests, manually collecting and analyzing data to refine processes. Morris Chang, also at TI, applied his methodical approach to dramatically improve manufacturing 'yields' for IBM computer transistors from near zero to 25 percent within months, demonstrating the critical role of engineering intuition and systematic process optimization in making chips viable.

While military and space programs provided the initial 'liftoff,' Bob Noyce envisioned and actively cultivated a much larger civilian market for chips. He deliberately kept Fairchild's R&D priorities separate from military contracts, which he viewed as less competent. Noyce strategically slashed chip prices—from 120 dollars in 1961 to 15 dollars by 1962, and then from 20 dollars to 2 dollars in the mid-1960s, sometimes even below manufacturing cost—gambling on expanding the civilian market. This aggressive pricing, coupled with Gordon Moore's prediction of exponential growth in chip components (Moore's Law), rapidly expanded civilian computer sales. By 1968, the computer industry purchased as many chips as the military, with Fairchild supplying 80 percent of this market. This shift marked a pivot for Silicon Valley, reducing its reliance on defense and space contracts, but also highlighted the growing employee dissatisfaction at Fairchild over the lack of stock options, as the desire for financial reward became a potent driver of innovation and industry growth.