Notes on Chapter 22: Computer on a Chip
Intel’s spartan office building – Intel’s first office, a modest, boxy building furnished with leftover furniture from Union Carbide, stood in deliberate contrast to the elegant corporate spaces Noyce had known at Fairchild. The layout of cubicles—placing Noyce up front, Grove near the manufacturing floor, and Moore between—reflected both function and Intel’s egalitarian culture [Berlin-1].
First MOS memory chips– After many trials and setbacks, Intel’s team succeeded in producing its first functional MOS memory chips—a breakthrough that prompted a celebratory champagne toast in the cafeteria. Noyce, who was away and recovering from a skiing injury, learned of the milestone by phone from Moore [Berlin-1].
Noyce’s “90-10” principle and flying exploits – Noyce’s “90-10” principle promoted using quick, rudimentary tests to reach 90 percent of a solution in 10 percent of the time, accelerating innovation at Intel. Moore dubbed it the “principle of minimum information,” and Grove became its chief enforcer. During this period, Noyce also earned his pilot’s license and bought a Pegasus aircraft [Berlin-1].
Busicom project and Hoff’s insight – The Busicom project—a contract to develop calculator chips—prompted Ted Hoff to envision a more flexible architecture: a general-purpose microprocessor. Hoff’s perspective proved pivotal to Intel’s leap into microprocessors [Malone-2], [V-Hoff].
“Tyranny of numbers” in circuit design– As the era of large-scale integration (LSI) approached, a new “tyranny of numbers” emerged—not in wiring, but in the sheer engineering effort required to design ever more complex custom chips. Noyce and Hoff saw general-purpose, programmable chips as the key to overcoming this bottleneck. See [Reid, p. 175].
Noyce’s encouragement to Hoff– In a casual but fateful remark, Noyce told Hoff, “Why don’t you go ahead and pursue your ideas?” This open-minded encouragement was instrumental in advancing the microprocessor project. The factors behind Noyce’s decision to allow Hoff to explore this alternative approach are explored in [Berlin-1] and [Malone-1].
