A first chip is only the beginning
The earliest integrated circuits of 1958 and 1959 held only a few components. They proved the idea but were costly and limited. What turned them into a technology that changed daily life was the steady shrinking and combining of circuits that followed, along with a manufacturing method that could repeat the results reliably.
The planar foundation
The key manufacturing step was the planar process, developed by Jean Hoerni at Fairchild Semiconductor in 1959. Parts were formed in a flat layer of silicon beneath a protective oxide skin, and metal lines were laid over the top. Because the structure was flat and sealed, it could be made with photographic printing steps and was more stable than earlier designs. The first commercial integrated circuits from Fairchild, around 1961, were simple logic circuits made this way and sold at high prices.
Counting by scale
Industry people described progress in levels of integration. Small-scale integration, in the early to mid-1960s, meant a handful of gates or transistors on a chip. Medium-scale integration, in the late 1960s, meant tens to hundreds of gates, enough to hold a counter or a small register. Large-scale integration, from the early 1970s, reached thousands of components, and very-large-scale integration, in the late 1970s and after, meant tens of thousands and beyond. The boundaries between these categories were loose, and different sources give different cutoffs, but the trend is clear.
Demand from space and defense
Governments were important early customers. The Apollo program and military missile projects needed lightweight electronics and were prepared to pay. This demand supported production lines when ordinary commercial markets were still small. As manufacturers gained experience, their yields improved and prices fell steeply, which opened up civilian uses such as hearing aids, calculators and computers.
Memory and calculators
Chips were useful not only for logic but for memory. Semiconductor memory started to compete with magnetic cores in the late 1960s and early 1970s, and Intel’s early products were memory chips. At the same time, the calculator market created a hunger for integrated circuits. Pocket calculators, which spread around 1970 and 1971, needed several chips at first, and makers searched for ways to cut that count.
The Intel 4004
A Japanese calculator company, Busicom, asked Intel to design a set of chips for a family of calculators. Intel engineers, including Ted Hoff, Stan Mazor and Federico Faggin, with input from Busicom’s Masatoshi Shima, proposed a small general-purpose processor that could be programmed to do the job, instead of a design hard-wired for each product. The result was the 4004, announced in 1971. It was a 4-bit processor, meaning it handled four bits of data at a time, and was made up of a little over two thousand transistors.
Calling the 4004 the first microprocessor is common, though some caution is wise. Other designs for a processor on few chips existed in parallel, and a classified aircraft flight-control computer built with a small number of chips has been argued to predate it. Texas Instruments also worked on a single-chip design around the same time. The 4004 is best described as the first commercially available microprocessor.
From 4004 to 8080
Intel’s 8008 followed in 1972, an 8-bit processor, and the 8080 of 1974 brought more speed and became the heart of early hobbyist computers. Other companies competed with their own designs, including Motorola and MOS Technology, and the hobby computing movement turned these chips into complete machines. That is how a part meant for a calculator became the core of personal computers.
Why chips got smaller
Shrinking worked because smaller transistors switch faster and use less power, and more of them fit on each wafer. The industry also found that cost per transistor fell steadily as long as improvements continued. Photolithography improved, wafers grew in diameter and cleanroom control tightened. The steady doubling of transistor counts was a mixture of design skill, process engineering and investment, and nobody could guarantee it would continue.
Miniaturization beyond computers
Smaller chips allowed products that had not been practical before, including the quartz watch, pocket calculators and personal music devices. Later, chips went into cameras, appliances, cars and smartphones. The idea behind chip miniaturization, squeezing more function into less space at lower cost, became a model for other technologies, including sensors and medical devices such as microprocessor-controlled prosthetics.
Looking back
The story after the first chip is not about a single invention but about steady improvement and a few large steps, such as the planar process and the microprocessor. Each step relied on the one before and on a growing community of engineers and companies.