Andrew Booth

Booth was doing X-ray crystallography at Birkbeck College and had the field's standard complaint: the structure only emerges from a Fourier synthesis that takes weeks to grind out. In 1947 he travelled to Princeton, met von Neumann, and came home with the stored-program idea. The following year the ARC — Automatic Relay Computer — went into service, followed by the SEC in 1949 and then the APE(X)C. All of them were small, cheap and built to serve his own arithmetic rather than to impress a funding body, which was unusual in an era when computer projects were measured in tons.

Two things outlived the machines. One is drum storage: in 1947 Booth built a rotating magnetic drum a couple of inches across and demonstrated that a moving magnetic medium worked as working memory, years before it became standard. The other is Booth's multiplication algorithm of 1950, which exploits runs of ones in the multiplier to skip additions and is still sitting in silicon today. His assistant Kathleen Britten wrote the machines' assembly language and later married him; the APE(X)C was copied by the British Tabulating Machine Company and became the ancestor of ICL's line.