A 1971 story, not a 1967 one
The Intel 4004, announced in November 1971, is usually named as the first commercial microprocessor. It was not a 1967 event: no complete processor on a single chip existed that year. The sections below describe how the 4004 came about, who deserves credit for what, and why some historians prefer to say “first commercial” rather than simply “first”.
The Busicom order
In 1969 the Japanese calculator maker Busicom approached Intel, then a very young company founded in 1968, to produce a set of custom chips for a new line of printing calculators. Busicom’s own design called for about a dozen different chips, each with a specific function.
Marcian “Ted” Hoff, an Intel engineer assigned to the project, thought the design was too complicated and too costly for Intel to make. He proposed instead a simpler arrangement built around a small general-purpose processor that would be told what to do by a stored program, with the calculator’s features supplied by software. Stan Mazor, another Intel engineer, worked with Hoff on the architecture and helped refine the instruction set. Busicom’s representatives, including the engineer Masatoshi Shima, reviewed the design and agreed to proceed.
Faggin and the silicon
A design on paper is not a chip. Federico Faggin joined Intel in April 1970. He had earlier worked at Fairchild, where he had helped develop silicon-gate process technology, and he took charge of turning the architecture into working silicon. This was difficult work involving circuit design and layout, and it had to be done on a tight schedule. Shima worked with him on the logic design. The first working chips are generally reported to have appeared in early 1971.
The finished 4004 was a four-bit processor containing roughly 2,300 transistors. It was one of a family of four chips, including memory and input-output support, but the processor itself was the novelty. By the standards of later chips it was very slow, running at under a megahertz, but it did the job of a computer’s central processing unit, the part that was in 1967 spread across boards in a cabinet.
Intel buys back the rights
Busicom had paid for the development and held the rights to the design. As the calculator market grew more competitive and prices fell, Busicom was reportedly keen to renegotiate. Intel secured the right to sell the chip to other customers, and in November 1971 it advertised the family under the headline of a new era of integrated electronics. That decision made the 4004 a general product rather than a custom part. Faggin reportedly put his initials on the chip’s silicon.
The 8008 and what followed
Intel’s next processor, the 8008, was an eight-bit design announced in 1972, originating in work for a computer terminal maker. It was followed by the 8080 in 1974. These chips were the basis of the early hobbyist machines, and from there the microprocessor became the heart of personal computers such as the Apple II. The Intel 4004 itself was used mainly in calculators and similar devices.
Competing claims
The phrase “first microprocessor” is contested, for good reasons:
- A processor designed by Garrett AiResearch for the F-14 fighter’s air data computer was completed around 1970 but was classified, and became publicly known only in the late 1990s. Its designers and some historians argue it deserves recognition.
- Lee Boysel at Four-Phase Systems built a single-chip processor section around 1969 for use in the company’s own systems, though it was designed as one part of a larger machine.
- Texas Instruments and Intel both worked on chips for the Datapoint terminal in 1970 to 1971. Texas Instruments later received a patent related to the single-chip computer idea, and patent disputes in the following years were complicated.
The result is that the 4004 is best described as the first commercially available single-chip microprocessor, and that is a more careful claim than “the first”. Several teams were working towards similar goals at the same time, helped by the rising chip density that had been under way since the 1960s.
Why it mattered
Before the microprocessor, building a computer meant assembling a processor from many parts. After it, the processor was a component, one that could be bought for a small sum and adapted through software. That change opened the way to personal computers, video games, digital watches, engine controllers and much else. It is a clear example of a technical shift whose effects went well beyond its first use, a calculator.