The first stored program for a computer
Research and development of a practical electromechanical calculator was being carried out at a number of American laboratories by 1942, when John Mauchly of Philadelphia and a colleague, J. Presper Eckert, drew up a technical outline of a machine intended to perform the laborious calculations needed to calculate the trajectories of artillery shells and submitted it to the Ballistic Research Laboratory (BRL) of the Army. The government soon devoted its resources to developing what would become the computer, and organized a project supervised by J.G. Brained and appointing Lt. Herman H. Goldstine as technical liason for the BRL.
The fruits of this first essay into modern computing was the Electronic Numerical Integrator and Computer (ENIAC), which was dedicated on February 15, 1946. Weighing in at over 30 tons, 150 feet in length, and consuming over 19,000 vacuum tubes, ENIAC was a giant, faster than any previous computer, able to perform fourteen 10-digit multiplications in a mere second - 500 times faster than the best mechanical calculators of the day. Despite its speed, however, it was hampered by highly limited in its internal memory and the need to rewire the computer manually for each separate program it was to run. Mauchly and Eckert had adopted such an inflexible design for ENIAC as a matter of expediency, given the demands of the wartime military, but they recognized that the inefficiency could be lessened by adopting serial calculation rather than parallel, and thus a new project, the Electronic Discrete Variable Computer (EDVAC) was born. Enter John Von Neumann.
What most distinguished EDVAC from ENIAC is that it used stored programs, so that instructions did not have to be input repeatedly. Von Neumann, an ambitious and talented mathematician, had learned of the ENIAC project in August, 1944, during a chance conversation with Herman Goldstine while awaiting a train. Von Neumann had been working on the Manhattan Project to develop the atomic bomb, and immediately recognized that an electronic computer could help work through the necessary calculations. By November, he joined the Mauchly and Eckert team, who had already begun development of the stored program computer.
The conceptual framework of EDVAC was fully in place by 1946, but delays caused by the end of the war and the departure of workers for civilian work delayed its completion until 1952. Although Mauchly, Eckert, and Von Neumann all contributed to its success, Von Neumann's 1945, draft outlining the principles behind the stored-program computer garnered him the lion's share of the credit for formulating the new approach. The 23-page document depicted in this exhibit was typed and written out by Von Neumann in 1945, and was intended to demonstrate the potential of the EDVAC design. It took as its task a very basic function, sorting, rearranging and merging data. Perhaps it can be said that with this small program, Von Neumann rearranged the past, merging it with the future.
John Von Neumann's instructions were donated to the Library by Herman H. Goldstine, former Executive Officer of the APS.
|Currently on exhibit|