Microelectronics and the Art of Similitude

A two-page paper delivered by Douglas Engelbart on a Friday morning in February 1960, at a conference devoted to making electronic components smaller. It survives mostly as a citation. Its argument has outlived nearly everything else presented that week.

The argument

The electronics industry of 1960 treated miniaturization as a manufacturing problem. Engelbart asked a different question: what happens when shrinking something changes the rules it runs by?

Applying similitude — the engineering science of scale — he showed that shrinking a circuit alters resistance, heat behavior, and noise in ways that do not scale proportionally. Past a certain point you are not building the same machine, smaller. You are building a different machine that happens to resemble it. He warned that devices would need modification in ways that would defeat intuitions built at normal size.

The half that got kept

Gordon Moore, then at Fairchild Semiconductor, was in the audience. Five years later he published the projection that became Moore’s Law.

The industry adopted the reassuring half of Engelbart’s claim — that the numbers improve on schedule — and set sixty years of roadmaps by it. The other half, that each new scale retires the old assumptions and opens fresh possibilities, stayed in the room.

Credit, all the way down

Engelbart did not invent similitude. The paper draws on Glenn Murphy’s Similitude in Engineering (The Ronald Press, 1950), the dimensional-analysis text Engelbart was working from — a source behind the source, and worth naming.