Loading...
By 1985, Intel was bleeding. The company that had invented the DRAM memory chip in 1968 now watched Japanese firms like NEC, Hitachi, and Fujitsu flood the market with cheaper, higher-quality chips. Intel's memory division—once 100% of revenue—had shrunk to under 30%, losing money every quarter. CEO Gordon Moore and president Andy Grove agonized for over a year, trapped by the identity they'd built. Intel meant memory. Their 8,000 employees in memory fabs, their sales relationships, their R&D infrastructure—all of it screamed 'stay the course.' Then came the moment. Grove turned to Moore in his office and asked: 'If we got kicked out and the board brought in a new CEO, what would he do?' Moore answered without hesitation: 'He'd get us out of memories.' Grove stared back: 'Why shouldn't ...
Popular framing: Andy Grove was bold enough to bet the company on a hunch.
Structural analysis: Intel was trapped by sunk cost in DRAM — 8,000 employees, fabs, identity, customer relationships — even as the unit economics collapsed against Japanese competition. The inversion ("what would a new CEO do?") was a structural device that severed sunk-cost anchoring and forced an opportunity-cost frame: every dollar in memory was a dollar not in microprocessors. The pivot worked because the 8086 had already been chosen by IBM; the architecture for the next S-curve was prepaid.
Centering the Grove-Moore conversation as the causal event obscures the degree to which the outcome depended on exogenous factors (IBM's chip selection, Japanese industrial policy, the PC ecosystem's explosive growth) that were not under Intel's control. This matters because the popular framing generates a replicable 'lesson'—ask the new CEO question—when the actual lesson is about recognizing which structural forces are terminal versus which can be redirected, a much harder and less transferable skill.
