Science: Eniac

After consuming and partly digesting lobster bisque, filet mignon and half a dozen speeches, half a hundred of the nation's topflight scientists and mathematicians got up from the table in the University of Pennsylvania's Houston Hall. They trooped three blocks down the street to the Moore Electrical Engineering School, to witness the first public demonstration of the latest and greatest mechanical brain—a series of dials, 18,000 tubes and cabinets, occupying an entire room.

The huge gadget was known as the "electronic numerical integrator and computer." Its inventors—Dr. John W. Mauchly and J. Presper Eckert—called it "eniac." For their blue-ribbon audience, they demonstrated how eniac could compute the trajectory of a shell in less time than it would take the shell to reach its target. Thus, in a trice, eniac showed its superiority over all its predecessors.

No Moving Parts. Calculating machines, which are as old as the Chinese abacus, have been growing steadily more complicated. But until now, they had not caught up with science's need for their services.

Harvard University has long had a calculator which would do almost any mathematical job. But it was slow; its inner works were like those of an ordinary adding machine. M.I.T. had a faster gadget, but it, too, was largely mechanical. What was needed, said Mauchly and Eckert, was a calculator with no moving parts except the fast-flying electrons inside vacuum tubes. With Army help and money, they built one, in 30 months.

No Mystery. Eniac does nothing mysterious. It does not need to, for the most complicated mathematical problem can be broken down into addition, subtraction, multiplication and division. The trouble is that these familiar operations, involving very large numbers, must usually be repeated over & over thousands of times in all sorts of combinations.

That is where eniac shines. Its nimble electrons can add two numbers of ten digits in 1/5000th of a second. New data can be fed into it at any stage of the process. If necessary, it can "remember" numbers and hold them for future use. An elaborate system of controls makes eniac so flexible that its inventors have given up looking for problems it cannot handle. There is plenty of work ahead for eniac, its inventors say. In nearly every science and every branch of engineering, there are proved principles which have lain dormant for years because their use required too much calculation. Example: aircraft designers know how to predict air-drag theoretically, but the job takes so much figuring that they prefer to make scale models and test them—somewhat inaccurately—in expensive wind tunnels. In future, they may rely on eniac.