ENIAC

Moscow State University.

Faculty of Computational Mathematics and Cybernetics.

The ENIAC Story

Group № 113

Kuznetsov Nikita

2009, September

The ENIAC Story

The world's first electronic digital computer was developed by Army Ordnance to compute World War II ballistic firing tables.

The stimulus which initiated and sustained the effort that produced the ENIAC (electronic numerical integrator and computer)--the world's first electronic digital computer--was provided by the extraordinary demand of war to find the solution to a task of surpassing importance. To understand this achievement, which literally ushered in an entirely new era in this century of startling scientific accomplishments, it is necessary to go back to 1939.

As the year 1939 dawned on an apprehensive and fearful Europe, soon to realize the worst of its fears with the outbreak of the war on September 1st, the United States continued largely oblivious to the outside world and its impending fate. This obliviousness was in no way better exemplified than in the size and state of unreadiness of the U.S. Army.

Two decades of complete indifference toward military preparedness had witnessed its virtual elimination as a factor of any military consequence in the world. In that fateful year the total strength of the Regular Establishment of the Army was approximately 120,000 officers and men.

The part of this exceedingly small peacetime establishment which provided the principal scientific and logistic support was the Ordnance Department. This Department had the responsibility for the design, development, procurement, storage, and issue of all combat materiel and munitions for the Army. In 1939 it was staffed by a relative handful of officers and career civilian employees.

The only scientific facility then available to the Ordnance Department for carrying out experimentation with weapons was the Aberdeen Proving Ground in Maryland. This facility had been acquired at the beginning of World War I and had been heroically maintained during the disheartening interim period so that at the outbreak of World War II it was able single-handedly to perform the crucial task of testing all combat materiel during the critical period of mobilization of the American war effort.

One of the extraordinarily important tasks which devolved upon the proving ground was the preparation of firing and bombing tables for the Army which at that time, of course, included the Army Air Corps. This responsibility was carried out at the Ballistic Research Laboratory of the Ordnance Department at Aberdeen. Here also were obtained experimental data of high accuracy and precision, necessary to the computation of the firing and bombing tables.

What was the situation at the Ballistic Research Laboratory on the eve of World War II? Its computing group comprised just a handful of civilian employees of the Ordnance Department. These individuals were well trained and highly skilled in the conventional methods of computation of firing and bombing tables. Available to this group at that time was one important calculating device other than standard desk calculators--this was the Bush differential analyzer.

This analogue device, or continuous variable calculator, had been installed at the proving ground about five years previously under the direction of Major James Guion of the Ordnance Department, then head of the ballistic computations section of the proving ground.

The analyzer installed at Aberdeen had ten integrating units and provisions for two input and two output tables as well. But, despite its value as an important mechanical aid to computation, it had several severe limitations. Probably the most severe of these was the mechanical torque amplifier. This element of the analyzer sufficiently amplified the extremely small torque developed by the integrating units so as to permit its transmission and utilization elsewhere in the device to drive other elements including other integrators.

This torque amplifier, although simple in mechanical design, frequently failed toward the end of a long trajectory run with the loss of the preceding computation and an appreciable delay associated with its repair.

The officer in charge of ballistic computations at that time was Lieutenant P. N. Gillon, Ordnance Department, who had just assumed responsibility for ballistic computations at the outbreak of the war in Europe. His immediate recognition of the immensity of the task that would devolve upon the Ordnance Department in the event of America's involvement in the war prompted him to seek both marked improvement in mechanical aids to computation and augmented facilities for their accomplishment.

It was, of course, known that the Moore School of Electrical Engineering of the University of Pennsylvania had a Bush differential analyzer of somewhat larger capacity than the one installed at Aberdeen. As a matter of fact, the one at the Moore School had fourteen integrating units. Therefore one of the first steps taken was the award to the University of Pennsylvania of a contract by the Ordnance Department for the utilization of this device.

Following the award of this contract, Lieutenant Gillon in his capacity as officer in charge of ballistic computations conferred frequently with Dean Harold Pender, Professor J. G. Brainerd, and their associates at the Moore School with a view to effecting proper coordination of the computational work at Philadelphia and Aberdeen.

Fortunately, at this time there was a very talented group at the Moore School under the direction of Professor Brainerd and as a result of Lieutenant Gillon's discussions with the professor and his associates, Assistant Professor Weygand undertook to develop an electronic torque amplifier to replace the mechanical torque amplifiers on the Bush differential analyzers. This work was eminently successful and in a rather brief period of time.

In addition, photoelectric followers were developed by the Moore School group for both the input and output tables of the analyzer. As a result of these accomplishments the productive capacity of the analyzers at both the Moore School and at Aberdeen were enhanced by at least an order of magnitude.

During the same period of time the computational activities at Aberdeen were being expanded greatly, and the increase in staff included both military and civilian personnel. Among the former, shortly after America's entry into the war, one of the very important individuals in the ENIAC story came to duty at the proving ground. This was Lieutenant Herman H. Goldstine, a Reserve officer of the Ordnance Department.

Lieutenant Goldstine had received his doctorate in mathematics at the University of Chicago under Professor Bliss who had, himself, been one of the principal ballisticians at the proving ground during World War I.

Upon reporting to active duty at the proving ground, Lieutenant Goldstine was assigned to the Ballistics Research Laboratory as an assistant to Captain Gillon. In view of the increased importance of the activities in Philadelphia, which by this time included a training responsibility in the mathematics of ballistic computations, Captain Gillon requested that Lieutenant Goldstine be assigned to duty at the University of Pennsylvania as supervisor of the computational and training activities there.

In September 1942, Colonel Gillon was assigned to the Office of the Chief of Ordnance as deputy chief of the Service Branch of the Technical Division with the responsibility for the research activities of the Department, including those at the respective Ordnance facilities. This, of course, included the work performed in the field of ballistic computations.

This responsibility required frequent contact with the activities at the University of Pennsylvania, and as a result thereof in the early part of 1943 Captain Goldstine and Professor Brainerd brought to Colonel Gillon the outline of the technical concepts underlying the development of the ENIAC. This outline had been prepared at Captain Goldstine's request by Dr. John W. Mauchly and J. P. Eckert, Jr.

Colonel Gillon fully realized the formidable opposition that probably would be offered to the initiation and prosecution of a development of this sort, especially in view of the highly speculative character of its successful completion. He was convinced, however, of the importance of the need not only to ballistic computations but also to the research activities of the Ordnance Corps as well, and accordingly he undertook to obtain the necessary authorization for its initiation and assumed full responsibility for its support and supervision.

The original agreement between the United States of America and the trustees of the University of Pennsylvania, dated June 5, 1943, called for six months of "research and development of an electronic numerical integrator and computer and delivery of a report thereon." This initial contract committed $61,700 in U.S. Army Ordnance funds.

Nine supplements to this contract extended the work to 1946, increased the amount ultimately to a total of $486,804.22, assigned technical supervision to the Ballistic Research Laboratories, and called for the delivery of a working "pilot model," first to be operable at the University of Pennsylvania and then to be delivered to the Ballistic Research Laboratories at the Aberdeen Proving Ground.

From this point forward, the research staff and faculty of the Moore School under Dr. Pender undertook rigorous prosecution of the development pursuant to the terms of the Ordnance contract. The project was placed under the supervision of Professor Brainerd, with Mr. Eckert as chief engineer and Dr. Mauchly, who provided the original outline for this development, as principal consultant. Captain Goldstine, the resident supervisor for the Ordnance Department, not only exercised extraordinarily detailed and highly competent supervision for the Government but also contributed greatly to the mathematical side of this undertaking. As in all important undertakings which achieve important results, this was the work of many individuals.

The ENIAC was placed in operation at the Moore School, component by component, beginning with the cycling unit and an accumulator in June 1944. This was followed in rapid succession by the initiating unit and function tables in September 1945 and the divider and square-root unit in October 1945. Final assembly took place during the fall of 1945.

By today's standards for electronic computers the ENIAC was a grotesque monster. Its thirty separate units, plus power supply and forced-air cooling, weighed over thirty tons. Its 19,000 vacuum tubes, 1,500 relays, and hundreds of thousands of resistors, capacitors, and inductors consumed almost 200 kilowatts of electrical power.

But ENIAC was the prototype from which most other modern computers evolved. It embodied almost all the components and concepts of today's high- speed, electronic digital computers. Its designers conceived what has now become standard circuitry such as the gate (logical "and" element), buffer (logical "or" element) and used a modified Eccles-Jordan flip-flop as a logical, high-speed storage-and-control device. The machine's counters and accumulators, with more sophisticated innovations, were made up of combinations of these basic elements.

ENIAC could discriminate the sign of a number, compare quantities for equality, add, subtract, multiply, divide, and extract square roots. ENIAC stored a maximum of twenty 10-digit decimal numbers. Its accumulators combined the functions of an adding machine and storage unit. No central memory unit existed, per se. Storage was localized within the functioning units of the computer.

The primary aim of the designers was to achieve speed by making ENIAC as all-electronic as possible. The only mechanical elements in the final product were actually external to the calculator itself. These were an IBM card reader for input, a card punch for output, and the 1,500 associated relays.

Another design objective was to make the electronics simple and reliable. This goal was achieved by utilizing vacuum tubes in a minimum of basic circuit combinations. To ensure reliable operation, circuits were constructed to rigidly tested standard components which were operated at current, voltage, and power levels below their normal ratings.

Accuracy of computation was assured by designing the basic circuits to work independently of the variable tolerances of their components. Numbers were not represented by electrical quantities which could be affected by changes in tolerance but only by the presence or absence of dynamic pulses.