1930s Berlin

IBM specialized in selling tabulating machinery to help businesses and governments manage their data processing. All they did was count punched cards. The problems of automating the mathematics that Babbage had been attacking a century before had still not been cracked. That did not mean that the problems had gone away. In fact if anything they had become worse. In the early 1930s, a young German called Konrad Zuse was studying to become a civil engineer at a technical university in Berlin and was fed up plowing though endless simultaneous equations for his stress calculations. He resolved to develop a device to help with those calculations and started work developing the machine in his parents’ living room. By 1938, his first machine, known as the Z1, finally worked – sort of. The machine could generally be coaxed, one way or another, to produce the required answer but reliable it was not. It was a pure mechanical computer like that proposed by Babbage – unlike Babbage’s, it was based on the binary system which kept the mechanisms relatively simple. It should be said that Zuse was not aware of the work of Babbage or indeed anyone else in the field. He came up with the ideas on his own in order to save him doing boring repetitive calculations.

The Zuse Z1 (linked from www.bnv-gz.de)

The ‘success’ of the Z1 spurred Zuse on to replace the less reliable parts of his design with alternative solutions but in 1939, war was declared and Zuse was called up to the German military. He was sent to the infantry but was later re-assigned as a stress analyst on flying bomb development which left him some spare time to continue with his work on calculating machines. Working when he could, he completed the Z3 with the help of similarly-minded friends and in 1941 the Z3, the first functioning program-controlled calculating machine, was demonstrated to the aircraft construction authorities. They were impressed but simultaneously underwhelmed: this device was not going to win the war. There was no obvious point in improving it or replicating it for pure military reasons. Nevertheless, the Z3 was used by the German Aircraft Research Institute in solving a number of problems, most notably that of ‘wing flutter’, but the research institute was not a high enough priority to allow Zuse to be released from his day job of stress analysis.

The Z3 he demonstrated was an ‘electro-mechanical’ device based on relays. Relays are similar, in a sense, to a domestic door-bell where a plunger is moved by an electro-magnet to hit a chime. In a relay the electrical contacts are used to move a switch. A single relay on its own is an extremely dumb device. In order to make a machine that can perform a complex task, many relays have to be wired together so that they can interact. The Z3 contained around 2600 interconnected relays, each wired to make decisions based on the settings of others.

A replica of the Zuse Z3 (linked from www.bnv-gz.de)

The drawback of using relays in calculating machines is that they are restricted by Newton’s laws of motion. It takes time for the contacts to move between ‘on’ and ‘off’ states and performing one calculation involves many, many such electro-mechanical decisions. Users of relay based computers had better have a good book to hand and be prepared to wait. The Z3 could only perform 3 additions per second or one multiplication every 5 seconds.

In 1942, work started on the Z4, an improved Z3. Work was hampered by the lack of resources due to war. High quality paper was in short supply so the program input tapes were made from used 35 mm celluloid film, presumably from movies no-one wanted to watch. In the closing weeks of the war, as the allies closed on Berlin, the incomplete machine was loaded onto a railway truck and began an Odyssey to safety in the south. Tension was high when the British and Americans finally sought out the then mysterious ‘V4’ (V stood for experimental). They were concerned that this was the big brother to the unmanned ‘Vengeance’ flying bombs, the V1 and V2, that caused devastation as they rained down on London during the late stages of the war. There was universal relief when they discovered it was nothing but a seemingly useless mass of electrical components.

Previous page: Early IBM
Next page: Harvard Mark I