Stanley Mazor was born in Chicago on 22 October 1941, which makes him 82 today. He is one of the four people credited with the invention of the microprocessor — the idea, now so obvious that it is hard to see as an idea, that an entire computer’s processing could be built on a single chip of silicon — and he is the one of the four who came to it as a programmer rather than an engineer.
From Fairchild to Intel
Mazor studied mathematics at San Francisco State University and joined Fairchild Semiconductor in 1964 as a programmer, moving into its research group to work on the Symbol computer, an ambitious machine that tried to put a high-level language into hardware. In 1969 he followed Fairchild’s founders to their new company, Intel, which was then a memory-chip maker with no interest in processors.
The calculator that became a computer
That year a Japanese company, Busicom, ordered a set of twelve custom chips for a desktop calculator. Intel’s Ted Hoff proposed doing it differently: instead of twelve chips each wired for one function, a single general-purpose processor chip that would run a program, with the calculator’s functions written in software. Mazor worked with Hoff on the architecture and wrote much of the instruction set and the software that proved it; Federico Faggin, who joined in 1970, did the circuit design that made it buildable, with Busicom’s Masatoshi Shima. The result, the Intel 4004, was delivered in 1971: 2,300 transistors, 4-bit, running at 740 kHz, with roughly the power of ENIAC in a chip the size of a fingernail. Busicom owned the design. Intel, seeing what it had, bought the rights back for $60,000 and, in November 1971, advertised the 4004 as “a new era of integrated electronics”. Every processor since — the 8080, the x86 line, ARM, the chip in your phone — descends from that decision.
The teacher
Mazor moved to Intel’s technical training group in 1977 and spent the rest of his career largely as a teacher: at Stanford, the University of Santa Clara, KTH in Stockholm and Stellenbosch University in South Africa. He joined Silicon Compiler Systems in 1984 and Synopsys from 1988 to 1994, where he co-wrote the standard textbooks on hardware description languages — the languages in which chips are now designed. In 1995 the IEEE invited him to write the history of the microcomputer for its proceedings; he was awarded the Kyoto Prize with Hoff, Faggin and Shima in 1997 and the National Medal of Technology in 2010.
The footnote
Bangladesh has no chip industry, but it has, increasingly, chip designers: several hundred engineers in Dhaka now write the hardware description code — in Verilog and VHDL, the languages Mazor’s books teach — for design houses in California and Taiwan, and the government has begun talking about a semiconductor design sector. Every one of them is working in a discipline that a programmer at a memory company helped invent by asking why a calculator needed twelve chips.
