Diff — ENIAC

**ENIAC** — the **Electronic Numerical Integrator and Computer** — was the first programmable, electronic, general-purpose digital computer. Designed by **John Mauchly** and **J. Presper Eckert** at the University of Pennsylvania's Moore School of Electrical Engineering and built for the U.S. Army between 1943 and 1945, it was publicly unveiled in Philadelphia in February 1946 and remained in service until October 1955. Roughly a thousand times faster than the electromechanical machines that preceded it, ENIAC is widely regarded as the beginning of the computer age.[[cite: Haigh, Priestley & Rope, [*ENIAC in Action*](https://mitpress.mit.edu/9780262535175/eniac-in-action/) (MIT Press, 2016)]][[cite: McCartney, *ENIAC: The Triumphs and Tragedies of the World's First Computer* (1999)]]

Its historical significance is doubled by who programmed it. ENIAC's first programming staff were six women — [[Kay McNulty]], [[Betty Jennings]], [[Betty Snyder]], [[Marlyn Wescoff]], [[Fran Bilas]], and [[Ruth Lichterman]] — who invented the discipline of programming in the act of preparing the machine for its work.[[cite: Kleiman, [*Proving Ground*](https://www.hachettebookgroup.com/titles/kathy-kleiman/proving-ground/9781538718285/) (2022)]]

ENIAC grew out of two problems that met at the Moore School in the early 1940s. The first was the Army's: the Ballistic Research Laboratory at Aberdeen Proving Ground could not produce artillery firing tables fast enough. Each table was a heavy computation performed by human "computers" — overwhelmingly women — working with desk calculators, and the war's appetite for new guns and shells made the backlog permanent.[[cite: Haigh, Priestley & Rope, [*ENIAC in Action*](https://mitpress.mit.edu/9780262535175/eniac-in-action/) (2016)]]

The second was Mauchly's. A physicist at Ursinus College who had built analog devices to model weather statistics, Mauchly had become convinced that digital electronic counting — vacuum tubes operating as switches — could outperform every other approach to computation. He arrived at the Moore School in 1941 as an instructor in its wartime engineering program, where he met the graduate student assigned to assist him: **J. Presper Eckert**, a brilliant circuit designer only twenty-two years old. In August 1942 Mauchly circulated a memorandum, *The Use of High Speed Vacuum Tube Devices for Calculating*, proposing an electronic digital calculator; in April 1943 the Army signed a contract — **Project PX** — for its construction, supervised on the Army side by Lieutenant **Herman Goldstine**, whose wife **Adele Goldstine** would write the machine's only manual and train its first programmers.[[cite: McCartney, *ENIAC* (1999); Goldstine, [*The Computer from Pascal to von Neumann*](https://press.princeton.edu/books/paperback/9780691023670/the-computer-from-pascal-to-von-neumann) (1972)]]

ENIAC was a **decimal** machine, not a binary one: each of its digits lived on a ten-position ring counter, and signed numbers were handled by complement arithmetic. Its core was twenty **accumulators** — units that could each store one ten-digit number and add or subtract at electronic speed. Around them sat specialized units: a high-speed **multiplier**, a **divider/square-rooter**, three **function tables** (banks of rotary switches holding up to 1,200 constants each), a **master programmer** for sequencing, an **initiating unit**, a **cycling unit** supplying the timing pulses, and punched-card input/output — an IBM card reader feeding a **constant transmitter**, and a card punch for results.[[cite: Haigh, Priestley & Rope, [*ENIAC in Action*](https://mitpress.mit.edu/9780262535175/eniac-in-action/) (2016); Adele Goldstine, *Complete Report of the ENIAC* (1946)]]

There was no program store in the modern sense. A computation was realized by *physically configuring the machine*: plugging patch cables between units on plugboards, setting switches, and arranging the accumulators and function tables into a plan. "Programming" ENIAC for a new problem could take days of rewiring — but the machine then executed in milliseconds what took human computers days.[[cite: Kleiman, [*Proving Ground*](https://www.hachettebookgroup.com/titles/kathy-kleiman/proving-ground/9781538718285/) (2022)]]

- Tube failure on a cycle of days: reliability, not speed, was the limiting engineering problem, and Eckert's conservative derating of components is what made continuous operation possible[[cite: [ENIAC in Action](https://mitpress.mit.edu/9780262535175/eniac-in-action/) (2016)]]

The Army selected six of its human computers to program the machine: [[Kay McNulty]], [[Betty Jennings]], [[Betty Snyder]], [[Marlyn Wescoff]], [[Fran Bilas]], and [[Ruth Lichterman]] — the **ENIAC six**, the first programmers. Trained by Adele Goldstine and given only the engineers' block diagrams, they taught themselves the machine and then taught it to compute — inventing flowcharts, subroutines, verification practice, and debugging as they went. Their full story is told at [[The first programmers]].[[cite: Kleiman, [*Proving Ground*](https://www.hachettebookgroup.com/titles/kathy-kleiman/proving-ground/9781538718285/) (2022)]]

ENIAC's first real computation was not artillery at all. In late 1945, before the public unveiling, the machine ran a classified feasibility calculation for **Los Alamos** — hydrodynamics work related to thermonuclear weapon research, associated with John von Neumann's circle. Firing-table work followed immediately afterward, at rates that made the corps' old backlog look medieval.[[cite: Haigh, Priestley & Rope, [*ENIAC in Action*](https://mitpress.mit.edu/9780262535175/eniac-in-action/) (2016)]]

On **February 14–15, 1946**, the Army unveiled ENIAC to the press at the Moore School. The showpiece was a sixty-second trajectory computed faster than the shell would have flown it; newspapers christened it the "Giant Brain." The six women who programmed it went unrecorded in the coverage — in some accounts introduced to reporters as models.[[cite: Kleiman, [*Proving Ground*](https://www.hachettebookgroup.com/titles/kathy-kleiman/proving-ground/9781538718285/) (2022); contemporary press coverage]]

In 1947 ENIAC was dismantled, moved to **Aberdeen Proving Ground, Maryland**, and rebuilt in Building 328, where it ran for eight more years. In 1948 it was converted to a **stored-program** mode of operation — the function tables repurposed as read-only memory holding instruction sequences — which removed the plugboard rewiring for most work and made it, by construction, one of the earliest machines to run stored programs. It continued in continuous service on ballistics, weather, and scientific computation until it was switched off for the last time on **October 2, 1955**.[[cite: Haigh, Priestley & Rope, [*ENIAC in Action*](https://mitpress.mit.edu/9780262535175/eniac-in-action/) (2016)]]

ENIAC's conceptual heir was **EDVAC**, the stored-program machine Mauchly, Eckert, and von Neumann's circle designed next — the famous 1945 draft report that spread the stored-program architecture worldwide. ENIAC itself is the ancestor by practice rather than by blueprint: it proved that electronic digital computation worked at scale, and it trained the first programmers, who carried the discipline into UNIVAC, COBOL, and the commercial industry.[[cite: Haigh, Priestley & Rope, [*ENIAC in Action*](https://mitpress.mit.edu/9780262535175/eniac-in-action/) (2016)]]

Physical pieces survive in several places: panels at the University of Pennsylvania and the Smithsonian's National Museum of American History, a functioning-era accumulator and units at Aberdeen, and fragments in collections worldwide. In 1996–97, for its fiftieth anniversary, University of Pennsylvania students re-implemented the entire machine on a single integrated circuit — **ENIAC-on-a-Chip**, compressing thirty tons of 1946 hardware into a die a few millimeters across. It holds IEEE Milestone status as the machine that began the era this site is built on.[[cite: [IEEE Milestone citation, ENIAC](https://ethw.org/ENIAC); University of Pennsylvania ENIAC-on-a-Chip project]]