Helmut Schreyer: The Missed Electronic Age
Abstract
Everyone who knows German computing history knows Konrad Zuse. Far fewer know the friend standing next to him: Helmut Schreyer (1912–1984), the telecommunications engineer who understood (years before ENIAC) that vacuum tubes could compute thousands of times faster than relays. Schreyer built working tube-logic prototypes in Berlin, proposed a 2,000-tube electronic computer, and was turned down twice: first by academics who called it fantasy, then by a wartime state that expected victory before the machine could be finished. He is one of computing history’s great “what if” figures, proof that a technological revolution can be stopped not by physics but by a political misjudgment.
The Friend with the Tubes
Helmut Schreyer was born on July 4, 1912, in Delitzsch, Saxony. From 1934 he studied electrical and telecommunications engineering at the Technische Hochschule Berlin-Charlottenburg, taking his diploma in 1938. In 1935 he met Konrad Zuse, and the two became close friends and collaborators; Schreyer was by most accounts the practical joker to Zuse’s obsessive builder, and he spent evenings helping assemble Zuse’s machines in the Zuse family apartment.
The division of labor between the two men shaped both their legacies. Zuse, the civil engineer, designed the logical architecture of his machines and chose electromechanical relays as the switching element for the Z3, proven telephone-exchange technology, slow but dependable. Schreyer, the telecommunications engineer, advised him on that relay technology. But Schreyer’s own conviction ran further: the same logical operations could be performed by vacuum tubes, which switch electronically rather than mechanically. A relay needs milliseconds to open or close; a tube switches in microseconds. An electronic computer would not be marginally faster than a relay machine; it would be faster by a factor of a thousand or more.
“Phantasterei”: The First Rejection
In 1938, Schreyer and Zuse presented the idea of an electronic computing machine to a small academic circle at the Technische Hochschule. When asked how many vacuum tubes such a machine would need, they answered: about 2,000 tubes, plus several thousand other components. The audience was incredulous. The most complex tube circuits then in existence (radio transmitters, early television experiments) contained no more than a few hundred tubes, and the power consumption and failure rates of a 2,000-tube machine struck the assembled experts as absurd. The proposal was dismissed as Phantasterei, a pipe dream.
The skepticism was not unreasonable by the standards of the day. It was simply wrong. Less than a decade later, ENIAC would run with about 17,000 tubes, and Tommy Flowers (another telephone engineer who trusted tubes) would build the 1,500-tube Colossus in wartime Britain against exactly the same institutional disbelief.
The Second Rejection: “The War Will Be Over First”
Schreyer did not stop at talk. In 1939 he submitted a plan to the German authorities for a full-scale electronic computer, the 2,000-tube machine, pitched among other things for aircraft-defense calculations. The military declined to fund it. The stated reasoning has become one of the bitter ironies of computing history: development would take roughly two years or more, and the war was expected to be over before the machine could be completed.
He built anyway. He made tube logic the subject of his doctoral work; his 1941 dissertation was titled Das Röhrenrelais und seine Schaltungstechnik (“The Tube Relay and Its Circuit Technology”) and set out experimental tube circuits implementing the same logical operations Zuse’s relays performed. By 1942 he had built an experimental machine with about 100 vacuum tubes; it was lost at the end of the war. In 1944 he built an electronic circuit for converting decimal numbers to binary.
Germany, which in 1941 possessed the world’s first working program-controlled computer and an engineer who knew how to make it electronic, chose not to take the step. The United States took it instead: ENIAC was funded in 1943 precisely because the U.S. Army was willing to bet on a long war.
Dead End: What Was Actually Lost
The Schreyer episode is a textbook case of how a technological revolution can be aborted by forecasting rather than physics.
- The technology worked. Schreyer’s prototypes functioned; his dissertation laid out a workable tube-logic circuit technique. Colossus (1944) and ENIAC (1945) proved within a few years that large tube machines were entirely feasible.
- The architecture existed. Zuse’s Z3 already had the logical design of a program-controlled binary computer. Marrying Zuse’s architecture to Schreyer’s electronics was precisely the combination that defined the postwar computer, but it happened in Philadelphia and Manchester, not Berlin.
- The decision was political, not technical. The rejection rested on a strategic assumption (a short war) that proved false. By the time the assumption collapsed, Germany’s research infrastructure was collapsing with it.
It is worth resisting the counterfactual’s most dramatic form: a German electronic computer would not have changed the war’s outcome, and Schreyer’s machine was proposed for calculation, not code-breaking. But for the history of computing, the loss was real. German computing, which had led the world in 1941, spent the postwar decades catching up (see Germany’s Computing Pioneers), and the electronic computer entered history as an Anglo-American invention.
Epilogue in Brazil
Schreyer had obtained a Brazilian passport in the closing days of the war, and he emigrated in 1949. In Rio de Janeiro he headed the telecommunications laboratory of the Brazilian postal service and taught as a professor at the military engineering institute (IME) and at PUC-Rio. There he closed the circle: with his students he worked on Lourinha, the first computer designed and assembled in Brazil. The man whose electronic computer Germany declined to build helped found the computing tradition of another continent. He died in Rio de Janeiro on December 12, 1984.