
Ted Hoff
Born in Rochester, New York, with a Stanford doctorate in electrical engineering. Joined Intel in 1968. Saw that one programmable processor could replace a set of custom chips and drew up the 4004’s architecture. Named Intel’s first Fellow in 1980.
Intel, Santa Clara, California · 15 November 1971 · Hoff, Faggin, Mazor and Noyce
A New Era of Integrated Electronics
In 1969 a computer’s processor filled whole circuit boards, and every new calculator needed its own custom chips. Intel was a year old and made memory. When a Japanese calculator maker ordered a set of chips, three Intel engineers proposed something bigger: one general-purpose processor that software could turn into anything. Two years later it was on sale.
A calculator order
Robert Noyce and Gordon Moore founded Intel in July 1968 to make memory chips. Nine months later Busicom, a Japanese calculator maker, asks Intel to build the chips for a new family of desktop calculators. In June three Busicom engineers arrive with their design, Masatoshi Shima among them. Each chip in it does one fixed job. Ted Hoff, Intel’s manager of applications research, is assigned to help them.
One chip that can do anything
Hoff fears the design is too complex for little Intel to build at Busicom’s price. He takes the worry to Noyce, who tells him to find a simpler way. Hoff sees one. A general-purpose processor that reads its program from memory could do all the work. Change the program, and the same chip runs a calculator, a cash register or a traffic light. Four bits at a time is enough: one decimal digit.
Robert Noyce to Ted Hoff, 1969, as Hoff recalled itWell, if you have any ideas to simplify the design, go ahead and pursue them.
Four chips
Stan Mazor joins Intel from Fairchild in September and works out the plan with Hoff. With Shima they answer Busicom’s objections one by one. Mazor writes a short interpreter to handle the calculator’s complex commands, and Shima reworks Busicom’s calculator logic around the new plan. In October Busicom’s executives hear the plan and accept it: four chips. One holds the program, one the data, one handles input and output, and one is the processor.
Faggin takes over
Neither Hoff nor Mazor designs chips, and the project has stalled. Intel hires Federico Faggin from Fairchild, where in 1968 he developed the silicon-gate process Intel builds its chips with. Days later Shima arrives from Japan and finds almost nothing done. Faggin brings two of his own inventions, the buried contact and the bootstrap load, to squeeze the processor onto one chip. He works 70- and 80-hour weeks, and designs all four chips in eight months.
Federico Faggin, on first seeing the scheduleMy jaw dropped: I had less than six months to design four chips, one of which, the CPU, was at the boundary of what was possible.
Buying back the rights
Busicom owns the design, and it is in trouble. It wants a lower price. Faggin urges Noyce to give it one in exchange for the right to sell the chips to everyone else, and Hoff tells marketing the same. Noyce makes the deal. Intel returns Busicom’s $60,000 and cuts the price, and Busicom keeps its rights only for calculators. Faggin builds a chip tester run by a 4004 to show what else it can do.
The team

Born in Vicenza, Italy, with a physics doctorate from Padua. Came to California in 1968 and developed silicon-gate chips at Fairchild. Designed the 4004 and signed it, then led the 8008 and 8080. Founded Zilog. American citizen since 1978.

Born in Chicago and raised in Oakland, California. Learned computing at San Francisco State and designed computers at Fairchild. Joined Intel in September 1969 and worked out the 4004’s architecture and instruction set with Hoff and Shima.

Co-inventor of the integrated circuit. Founded Intel with Gordon Moore in 1968. Told Hoff to pursue the simpler design, then traded back Busicom’s $60,000 for the right to sell the 4004 to the world.