John von Neumann

“It would appear that we have reached the limits of what is possible to achieve with computer technology, although one should be careful with such statements, as they tend to sound pretty silly in 5 years.” (John von Neumann, 1949)

A world-renowned mathematician, von Neumann has also created a pioneering place for himself in the annals of computing, as the architecture of all modern computers are based on the stored program principle, the so-called “von Neumann architecture”.

Von Neumann was born as János Lajos Neumann on December 23, 1903, in Budapest, Hungary. He was the first son of a prominent Jewish banker, Max Neumann. The family employed German and French governesses and the Neumann children – John was the oldest of three – learned their foreign languages at an early age from their caretakers and private teachers. In addition his father made sure his children learned the classics, so they could even converse in Greek at the dinner table. As a child von Neumann was called Jancsi – the diminutive form of János, or Johnny. He was a child prodigy and even at the early age of six he showed remarkable feats of memorization. He could memorize a whole page’s column from a telephone directory and could tell you which name and address belonged to which number.

In 1913 Max Neumann purchased his family a title but he never used the Hungarian “Margittai” – the equivalent of the German-Austrian “von” – in the name himself. His son however changed his name to János von Neumann.

At age eight von Neumann entered the Lutheran Fisori Gymnasium in Budapest where his teachers quickly discovered his mathematical talents and extra maths tuition was organized for the young boy. After completing his education at the gymnasium in 1921, he published his first mathematical paper the next year. Despite his obvious talent for mathematics, his father didn’t think maths would be financially rewarding enough and he managed to convince his son to change to a potentially more rewarding subject, so von Neumann started studying chemistry. Despite being Jewish, he was also accepted at the University of Budapest to study mathematics, but he skipped all classes and just turned up for the examinations. Simultaneously he started his study in chemistry at the University of Berlin in 1921. In 1923 he changed from Berlin to the Technical Institute of Zürich where he received his diploma in chemistry in 1926. The same year he graduated from the University of Budapest with a Ph.D. in Mathematics with a thesis on set theory. Following this he started lecturing mathematics at the universities of Berlin, Hamburg and Göttingen. His lectures, combined with a number of papers on complex mathematical subjects, soon established the young von Neumann as the new genius in the field of mathematics.

                                                                 with explanatory text on reverse

In 1930 von Neumann was invited to Princeton University (New Jersey) where in 1933 he was appointed professor in Mathematics. In 1933 Princeton founded the Institute for Advanced Studies (IAS) and von Neumann was one of the six original mathematics professors selected for the prestigious position – Albert Einstein was another member of this select group. Von Neumann would keep this position till his death. He still held his academic positions in Germany but when in 1933 the Nazis under Hitler came to power, he resigned all of them and decided, together with his mother and brothers, to move permanently to the United States where he foresaw better job opportunities for Jewish academics. He anglicized his name to John von Neumann and four years later he applied for American citizenship.

In 1936 he met Alan Turing (both are shown on the stamp on the left) who had become a graduate student at Princeton’s Department of Mathematics for his dissertation. Two years earlier Turing had published his “On Computable Numbers with an Application to the Entscheidungs-problem” describing his ideas about the universal Turing machine and its logical design. It is likely that von Neumann was aware of Turing’s paper and the ideas expressed in it as he invited Turing to become his assistant. However, Turing declined the offer and returned in 1938 to England where he would get involved with the code breakers of Bletchley Park and the development of the Colossus computer, which, incidentally, was operational long before the ENIAC was.

World War II

Von Neumann was a mathematical genius who did seminal work on quantum mechanics, set theory, game theory (a mathematical approach to the logic of conflict situations, used in economics, socio-political and military applications), logic and operator theory, the hydrodynamics of explosions and the theory of shocks, ballistics, statistics, the concepts and theory of automata, etc. His interest in shock waves got him invited to the Ballistics Research Laboratory at the Aberdeen Proving Ground (see above cover) in Maryland in 1937 where he became a member of the Scientific Advisory Commission. In 1943 he also joined the Manhattan Project at the Los Alamos Scientific Laboratory – the Manhattan Project team was responsible for the development and deployment of the first atomic bombs which ended World War II. In 1945 von Neumann was appointed manager of the Electronic Computer Project at Princeton, a position he would keep until his death.

The von Neumann Architecture stored program computer

Because of his work on ballistics at the Aberdeen Proving Ground von Neumann had a serious need for extra calculating power and he enquired what was available in that field. One day, by sheer coincidence, he met Herman Goldstine – a professor in Mathematics involved with the ENIAC project - at the local railway station at Aberdeen and heard about the development of the ENIAC (Electronic Numerator, Integrator and Computer) at the Moore School of Electrical Engineering at the University of Pennsylvania, Philadelphia. The Moore School’s ENIAC project was funded by the Ballistics Research Laboratory as they were in dire need of a faster way of calculating firing tables. The calculations for any ballistic trajectory for field or naval artillery were lengthy and involved lots of variables regarding distance, weight of the shell, wind speed and direction, humidity, gun elevation and many other factors. The people who did these calculations and composed the firing tables, were called computers. But as it could take up to 20 hours for a human computer to calculate a single ballistic trajectory, urgent help was required to speed up this process – in due course the ENIAC would be able to calculate a trajectory in just 30 seconds. The driving force at the Moore School's ENIAC project were J. Presper Eckert and John Mauchly, and it was to these two people that von Neumann introduced himself.

One of the problems Mauchly and Eckert were addressing, was the way to store a computer program. Earlier computers, like the Zuse computer, had used punched card input for this purpose, but Mauchly and Eckert soon realized, that the program had to be stored in the computer itself. If the computer was to execute a program at the speed of electronics, it had to be able to read the program at that speed as well. This could only be achieved if the program was stored internally in its memory, similarly to the way the data was held and read. Although today’s modern computers all use this principle, in the early days of computers these ideas were quite novel.

The designers of the ENIAC and its successor - the EDVAC (Electronic Discrete Variable Computer) - used group discussions to solve this and other problems and were now joined by von Neumann who participated enthusiastically. As not everyone was always present during these discussions it became important to document some of the discussions and their findings. Von Neumann volunteered for this role as he had a knack for documenting thoughts and expanding on them as he went along. He produced a document, called “First Draft of a Report on the EDVAC”, in which the architecture and design of a stored program computer was described in detail. The draft report did not yet contain a list of contributors or give credit to any particular idea or contribution as this would be added in its final version. Unfortunately, one member of the project team who received the draft, added a cover page listing von Neumann as the only author (of the draft) and subsequently copied this to others. It was a mistake resulting in confusion, misunderstanding and even acrimony. The draft was the first important document to describe the basic principles – later to be called the von Neumann principles – for an electronic stored program computer:

        ● application of the binary system, where bits – or binary digits – can only have the value “0” or “1”. He also proposed the bit as the measuring unit for computer memory.
        ● the computer should operate completely electronically
        ● the use of the central processing unit and an arithmetical unit
        ● storage and control of a program and data storage.

Coming from the highly respected and influential von Neumann and with his name on the front of the document, he inadvertently received the credits for its contents and principles. This type of architecture would be called the Von Neumann Architecture and would become the standard for virtually all mainframe and personal computers that followed.

Obviously Eckert, Mauchly and the other project members were a bit upset – to put it mildly – at seeing their names not mentioned and receiving no credit for their ideas. It resulted in many acrimonious arguments and fights about patent rights, ultimately resulting in Eckert and Mauchly resigning in March 1946 to set up their own company where they built the first UNIVAC computer. In 1950 their company would be taken over by Remington-Rand.

Note the misspelled surname

After the war, von Neumann went back to IAS at Princeton and proceeded to make significant contributions to the development of logic design and automata theory. He also continued to work on the development of electronic computers. Herman Goldstine was one of the few members of the original ENIAC project who joined him at IAS, as most other ENIAC staff members declined to work with him. The first IAS computer saw the light in 1952. It had a memory of 1024 words of 40 bits each, could do a simple addition in about 60 microseconds and contained 2,300 vacuum tubes – compare that to the ENIAC’s 17,468 vacuum tubes. After numerous enhancements, the machine was retired in 1960 and donated to the Smithsonian Institute in Washington, D.C.

In 1956 von Neumann was diagnosed with bone cancer, possibly as a result of too much exposure to radioactivity during his participation in the Manhattan Project. He died a painful death only a few months later on February 8, 1957. He was only 53.

Note that on the Hungarian stamps the family name
(Neumann), following Hungarian custom, is stated first.

Von Neumann has been honored in many ways – the Von Neumann crater on the moon has been named after him, Princeton University named its computer center the John von Neumann Computing Center, and in 1956 – shortly before his death – President Eisenhower awarded him the Presidential Medal of Freedom. In 1990 the Institute for Electrical and Electronics Engineers (IEEE) established its IEEE John von Neumann Medal “for outstanding achievements in computer-related science and technology.” Both Hungary and the United States have issued stamps to commemorate this computer legend and mathematical genius.

© Wobbe Vegter, 2008

[ Back ]