Intel quit the memory-chip business it was founded on to bet on microprocessors

Intel had been founded in 1968 as a memory-chip company, and DRAM was still central to its identity in the mid-1980s even as Japanese manufacturers priced it into a low-margin commodity Intel couldn't compete on. The company posted a $173 million net loss in 1986, according to its own 1987 annual report, one of what Intel later called 'two tough years.' Memory chips that had built the company's reputation were now the business dragging it down.

On October 10, 1985, CEO Andy Grove and Chairman Gordon Moore decided Intel would leave the DRAM business entirely, a moment Grove described using a now-famous test: he asked Moore what a new CEO brought in from outside would do, and when Moore said 'get out of memories,' Grove replied, 'why shouldn't you and I walk out the door, come back in, and do it ourselves?' Intel's own corporate history dates the decision to that day, with Grove later calling it 'the best business decision we ever made.' Intel redirected its capacity and engineering effort to the x86 microprocessor line instead.

The bet paid off within two years: net income reached $248 million in 1987, with revenue up 51% over 1986, as the 486 microprocessor family drove demand. By the early 1990s the shift had fully compounded — Intel's FY1993 10-K reports net revenues climbing from $4,779 million in 1991 to $8,782 million in 1993, and net income from $819 million to $2,295 million over the same span, attributing the growth explicitly to the market's shift toward higher-performance microprocessors.

Grove and Moore asked what an outside CEO would do, then exited DRAM to focus on x86; net income hit $2.30b by 1993.

Exiting a commodity business Intel could no longer win on price freed capacity and engineers for microprocessors, a market where design, not manufacturing cost, decided who won.

Framing the call as what an outside replacement CEO would do let Grove and Moore treat memory chips as a sunk identity rather than a business still worth defending.

Committing fully rather than hedging with a shrinking DRAM unit meant Intel's engineering resources went entirely toward the 386 and 486 lines that ended up funding the recovery.

When a founding product becomes a commodity you can't win on price, walking away and redirecting resources to your real advantage can beat defending a shrinking legacy business.

Intel's net income grew from $819 million in 1991 to $2,295 million in 1993, per its SEC filings, after posting a $173 million loss in 1986 during its last years in memory chips. The 1985 exit from DRAM became one of the most cited pivot decisions in Silicon Valley, discussed at length in Grove's own book Only the Paranoid Survive, and Intel's microprocessor line went on to dominate the PC market for decades after the decision.