No, Bill Gates Never Said '640K Ought to Be Enough for Anyone'

“640K ought to be enough for anyone” is repeated constantly as an example of a famously wrong tech prediction — the problem is there’s no verified evidence Bill Gates ever actually said it, and he has repeatedly and explicitly denied it on the record.

Where the quote actually first appears in print

The earliest known print attribution traces to an April 29, 1985 InfoWorld magazine article by James Fawcette — without any citation or sourced context for the quote. No recording, transcript, or contemporaneous interview has ever surfaced containing these words from Gates.

Gates’s own repeated denials

Gates has addressed this directly on multiple occasions: “I keep bumping into that silly quotation attributed to me that says 640K of memory is enough. There’s never a citation; the quotation just floats like a rumor, repeated again and again.” In a separate newspaper column, he wrote: “I’ve said some stupid things and some wrong things, but not that. No one involved in computers would ever say that a certain amount of memory is enough for all time.”

What Gates actually said, on the record, about the 640K limit

In a recorded 1989 speech, Gates gave a more measured and specific account: discussing the jump from 64K to 640K in 1981, he said, “I felt like I was providing enough freedom for 10 years. That is, a move from 64k to 640k felt like something that would last a great deal of time” — a claim about a roughly ten-year runway, not an unbounded, permanent ceiling. Gates has also noted the 640K limit itself was a consequence of processor architecture decisions, and that he had actually pushed to increase it.

Where the 640K barrier actually came from

The number itself was never a Microsoft promise to begin with. The IBM PC’s Intel 8088 could address one megabyte of memory through its 20-bit address bus, and IBM’s memory map — not MS-DOS — reserved the upper 384K of that space for video memory, the BIOS ROM, and expansion card ROMs, leaving 640K of conventional memory for programs. Attributions of the quote usually place it at a 1981 trade show, precisely when that map was being set, but no attendee, recording, or contemporaneous report has ever corroborated it.

What is documented is that the limit became the industry’s shared headache within a few years, and that an entire engineering subculture grew up around escaping it: bank-switched expanded memory under the LIM EMS specification, extended memory managed through XMS and HIMEM.SYS, and eventually DOS extenders that let protected-mode programs use megabytes while still booting from DOS. The HIMEM and EMM386 configuration battles that defined late-DOS-era troubleshooting are the practical legacy of that address map. None of that history requires the quote to be real — which is convenient, because nothing suggests it is.

Why the misattributed version stuck anyway

The mangled, absolutist “enough for anyone” phrasing is a much better story than Gates’s actual, hedged ten-year estimate — it fits a satisfying narrative of a tech titan making a laughably shortsighted permanent prediction, which is a more shareable anecdote than the real, more mundane claim about 1980s memory architecture roadmaps.

The lesson for citing famous tech quotes generally

A quote that “everyone knows” but that no one can trace to an actual recording, transcript, or contemporaneous, attributed source should be treated with real skepticism — the 640K quote is a clean example where checking the actual paper trail (an unsourced 1985 magazine mention, followed by direct, repeated denials from the person it’s attributed to) tells a meaningfully different story than the popular version.

Related:

Sources: