1965: Moore Predicts Chip Growth

On This Day in Business: April 19, 1965

On April 19, 1965, Electronics magazine published Gordon Moore's article predicting the rapid growth of components on integrated circuits. Moore, then director of research and development at Fairchild Semiconductor, observed that the number of components on the most economical chips had roughly doubled each year and expected the trend to continue for about a decade. He illustrated how increasing integration could make electronics cheaper and more capable, eventually supporting home computers, automatic controls, communications equipment, and other applications that still seemed futuristic. The forecast later became known as Moore's Law, though it was an observation and industry target rather than a physical law.

Semiconductor economics made the prediction powerful. Placing more functions on one chip could reduce the cost per component while improving speed and reliability, but only if engineers advanced lithography, materials, design, manufacturing precision, and yield. Each generation required expensive equipment and research. Suppliers, chipmakers, computer companies, and software developers began planning around a shared expectation of continuing improvement. Moore revised the pace in 1975 to approximately a doubling every two years, and related formulations often described performance or transistor density over similar intervals.

The expectation became partly self-reinforcing. Road maps helped coordinate investments across a complex supply chain, while intense competition pushed firms to meet the predicted curve. Intel, which Moore co-founded in 1968, became a major beneficiary and driver of the trend through memory and microprocessor products. Computing power moved into businesses, homes, vehicles, phones, and industrial systems as costs fell. At the same time, fabrication plants grew vastly more capital intensive, concentrating leading-edge production among a small number of companies and making global supply dependencies strategically important.

Moore's 1965 article shows how a credible forecast can organize an industry. It provided no guarantee, but it gave managers a common horizon for product planning and investment. Customers could anticipate better capabilities, and developers could design for hardware that did not yet exist. Physical and economic limits have slowed some traditional measures, yet innovation continues through specialized chips, new packaging, parallelism, and software. The business lesson is not that exponential progress is automatic. It is that sustained improvement requires coordinated learning, enormous reinvestment, and a market willing to convert each technical advance into useful demand. The forecast also shaped talent markets. Engineers chose specialties, universities built programs, and suppliers invested in techniques that appeared likely to matter for the next generation. Shared expectations lowered some coordination uncertainty while raising the stakes for companies that missed a cycle. Because progress compounded, one delayed process could damage several future products. Semiconductor strategy became a race of synchronized learning in which equipment, design, manufacturing, and customers all had to advance together.