First Draft of a Report on the EDVAC

The First Draft of a Report on the EDVAC is an important document written by John von Neumann. It was shared with others on June 30, 1945 by Herman Goldstine, who worked on a secret computer project called the ENIAC. This document was not finished, but it had 101 pages.

In this report, von Neumann described how a computer could work using a special idea called the stored-program concept. This idea later became known as the von Neumann architecture. However, some people think the name is not fair because von Neumann did not mention the other important people who helped create the ENIAC, John Mauchly and J. Presper Eckert.

History

John von Neumann wrote this report by hand while riding the train to Los Alamos, New Mexico and sent his notes to Philadelphia. Herman Goldstine had the report typed and made copies. Though the official date on the typed report was June 30, 24 copies were shared with key people involved in the EDVAC project five days earlier, on June 25. The report sparked so much interest that it was shared worldwide. Maurice Wilkes from Cambridge University said the report excited him so much that it inspired him to travel to the United States for the Moore School Lectures in the summer of 1946.

Synopsis

John von Neumann wrote about a design for a very fast digital computer. He split the computer into six main parts: a central part for arithmetic, a control part, memory, input, output, and slower external memory like punched cards or tapes.

The arithmetic part would handle basic math such as addition and multiplication. Numbers would be shown in binary notation, using 32 bits for each number. The design would use vacuum tubes to switch quickly, making the computer faster than older machines.

Von Neumann suggested using a single memory space for both numbers and instructions, which later became known as the von Neumann architecture. He thought the computer would need quite a bit of memory, estimating around 8,192 words of 32 bits each. He considered different ways to store this memory, like delay lines and iconoscope tubes. He also described different types of instructions the computer could follow, such as basic math operations and moving data between parts of the computer.

punched cards Teletype tape magnetic wire steel tape table look up interpolation biquadratic binary notation bit Claude Shannon Two's complement Vacuum tubes relays neuron AND gate Boolean logic vacuum tube oscillator crystal controlled Von Neumann architecture ordinary partial differential equations sorting probability experiments delay line iconoscope refreshed

Controversy

The sharing of this report caused strong disagreements among the EDVAC design team. Some believed the report was made public too early, which affected a later patent. Others felt that the ideas in the report came from earlier discussions at the University of Pennsylvania's Moore School of Electrical Engineering, and that von Neumann did not include everyone who helped develop these ideas. Because of this, von Neumann received all the credit for the report alone.