Atanasoff Berry Computer

John Vincent Atanasoff, a physics professor at Iowa State College, sketched the core idea of electronic binary computing on a napkin at a roadhouse in Illinois during the winter of 1937. Frustrated after years of fighting slow, error-prone mechanical calculators, he drove hundreds of miles from Ames with no destination in mind except to think, stopped at a bar across the state line, and over a drink worked out three ideas that would define the machine: base-2 binary arithmetic instead of decimal, a strict separation between memory and the logic that computes, and a way to store binary digits as electrical charge that could be refreshed before it leaked away. He began building the machine with graduate student Clifford Berry around 1939, and by 1942 the Atanasoff-Berry Computer, ABC for short, was working at Iowa State. It was the first machine in this history to compute with electronic binary logic — vacuum tubes rather than Konrad Zuse's sliding metal plates or Herman Hollerith's electromechanical relays, switching thousands of times faster than anything mechanical could manage. About 280 vacuum tubes handled the arithmetic. Its real signature, though, was regenerative capacitor memory: binary digits were stored as electrical charge on a rotating drum lined with capacitors, and because that charge naturally decayed, the machine had to periodically read each value and rewrite it before it faded past recognition. That refresh requirement is not a historical curiosity — it is the direct conceptual ancestor of the refresh cycle every DRAM chip performs today, the same basic problem of leaking charge solved with the same basic answer of reading and rewriting before the bit is lost. The ABC was not a general-purpose, programmable computer in the modern sense; it had no stored-program capability and was not Turing-complete. It was built for one job, solving systems of up to 29 simultaneous linear equations, and it did that job well. Its historical importance rests specifically on being first to prove electronic binary computation worked at all, not on general-purpose programmability, which was still a few years and a few more machines away. The ABC also has a quieter legal afterlife worth stating plainly: in 1973, a US federal court ruling in Honeywell v. Sperry Rand invalidated the ENIAC patent, finding in part that Atanasoff's earlier work had anticipated key elements of it, after ENIAC co-designer John Mauchly visited Atanasoff and examined the ABC in person in 1941. The ruling did not erase ENIAC's own importance, but it formally recognized Iowa State's machine as sitting earlier in the lineage than its patent holders had claimed. Where this history goes next is Britain, 1943: Bletchley Park's Colossus scaled vacuum-tube electronic computing from the ABC's roughly 280 tubes to somewhere between 1,500 and 2,400, and for the first time gave that electronic speed real wartime operational programmability, built specifically to break the Lorenz cipher used by the German high command.

Taken together, the Atanasoff-Berry Computer is less a complete machine in the modern sense than a proof that electronic binary logic worked at all, and a working demonstration of a memory-refresh problem that every DRAM chip built since has had to solve the same way Atanasoff solved it on a rotating capacitor drum in 1942.