Stibitz, George Robert

views updated

Stibitz, George Robert

(b. 30 April 1904 in York, Pennsylvania; d. 31 January 1995 in Hanover, New Hampshire), inventor of the modern digital computer.

In his early childhood, Stibitz’s family moved from York, Pennsylvania, to Ohio, where he attended Moraine Park, an experimental school in Dayton. He then went on to Denison University, where he received his Ph.B. degree in 1926. After graduation, Stibitz moved to Schenectady, New York, where he combined work at the General Electric research laboratories with pursuit of an M.S. degree from Union College, which he obtained in 1927. Stibitz then attended Cornell University, where he earned his Ph.D. degree in mathematical physics in 1930.

Work was not easy to obtain in 1930, at the outset of the Great Depression, but Stibitz succeeded in getting a job as a mathematical consultant at Bell Telephone Laboratories. The job gave him an opportunity to tinker, a lifelong preoccupation. Stibitz would ultimately hold thirty-eight patents in his own name, including one for a stereophonic organ, along with numerous others held in the name of Bell Labs. In 1937 he was asked to study the magneto-mechanics of telephone relays, the electromechanical switches used to connect phone circuits. His focus quickly moved from the relays themselves to the binary circuits that the relays created, and to the many possible arithmetic calculations that were reducible to binary form. Finally, Stibitz decided to construct his own binary adding machine, combining several relays with flashlight bulbs and connecting them all with metal strips cut from a tobacco can.

Stibitz’s Model K adding machine (so called because he built it at his kitchen table) used electromechanical relays as “gates.” If the relays for the integers 5 and 3 were activated together, for instance, they would open the gate of the relay for 8, their sum, thus lighting the appropriate bulbs. A replica of this first crude machine is displayed at the Smithsonian Institution. Over the next two years, working with Samuel B. Williams, Stibitz created a far more sophisticated machine called the Complex Number Calculator, which could perform all four basic arithmetic operations, using both real and complex numbers. This machine, containing 450 telephone relays and ten crossbar switches, could divide two eight-place complex numbers in about half a minute. It could both read input from a teletypewriter and print the output the same way. It first operated on 8 January 1940 and was installed at Bell Labs’ main office in New York City, where it was linked to three teletypewriter “terminals.” It took up about twenty cubic feet.

Later that year, Stibitz pulled off a public relations coup for his new machine. On 11 September 1940 the American Mathematical Society held a meeting at Dartmouth College in Hanover, New Hampshire, and Stibitz attended, bringing his teletypewriter. Telephoning New York City, Stibitz allowed the assembled mathematicians to pose arithmetic problems, which he then “downloaded” to his machine some 250 miles away. The answers printed out within seconds. This public demonstration of a working digital machine received wide publicity, and Bell Labs promptly renamed Stibitz’s “calculator” the Model 1 Relay Computer.

Stibitz’s demonstration did more than show how a digital computer could work; it displayed the principles of “remote entry” which would make the computer commercially viable. Particularly during the early decades of computing, when the huge machines took up entire rooms, the ability to access the mainframe (as it came to be called) through remote terminals was invaluable. Data could be entered at a laboratory, a workstation, or a job site; the calculations could be performed hundreds of miles away, at the home office; and the results could be displayed at a sales meeting, a production conference, or even at home.

During World War II Stibitz worked for the National Defense Research Committee (NDRC), joining with other early computer experts to help predict his infant specialty’s future course of development. As with many experts during an industry’s formative years, his predictions were wrong. Stibitz had gained tremendous success with computers based on electromechanical relays, whose inherent reliability made up for their relatively slow performance. He was understandably reluctant to back a new generation of computers based on speedy but unreliable electronic vacuum tubes. Indeed, until the commercial development of the transistor in the 1950s, electromechanical computers did retain a significant edge in reliability. But Stibitz and the rest of Subcommittee Z on High-Speed Computing (which grew out of the NDRC) failed to see the revolutionary potential of UNIVAC and other new designs, which hindsight would show to have laid the foundation of modern mainframe computing. The radical innovations that would make UNIVAC the first commercially successful mainframe computer, such as its use of high-speed magnetic tape rather than teletypewriter or paper tape for input and output, were dismissed as “details” in Subcommittee Z’s final report.

Disappointed by his failure to influence government policy and by his apparent inability to reach useful conclusions about the emerging new computer technologies, Stibitz established himself as a private consultant in Burlington, Vermont. However, he continued to tinker, and in 1954 he developed a prototype of the minicomputer—a full-function machine the size of a desk rather than a room—a decade before this kind of computer would become popular. But never again would he experience that thrill of the years 1938–1940, when a brilliant idea met its ideal moment of opportunity.

At the age of sixty, Stibitz joined the Dartmouth Medical School, where he worked on applications of computers to biomedicine. He was given the rank of professor in 1966 and became professor emeritus in 1972, continuing his work there until 1983. During his later years, when most people are enjoying retirement, Stibitz worked on such subjects as the anatomy of the brain and the movement of drugs through the body, and he developed a mathematical model of the capillary interface, in which arterial blood gives up its oxygenated hemoglobin and becomes venous.

Stibitz married Dorothea Lamson in 1 September 1930 and had two daughters. He died at the age of ninety at his home in Hanover and is buried in that city.

Stibitz, who cut an imposing figure with his tall stature and long, lean face, was not the only member of his family to achieve long life; he was survived by a brother and two sisters. He was granted a form of happiness gained by few people: he was able to spend his life doing something he loved, gaining success and admiration in the process. His computers may appear crude to modern eyes, but Stibitz’s electromechanical machines paved the way not only for the electronic computers that followed, but also—through his invention of remote job-entry—for the decentralized Internet-based culture of the modern world.

For more information on Stibitz, see Joel Shurkin, Engines of the Mind (1996). Stibitz’s life and achievements are also memorialized in the web sites of the AT&T Labs Research History, administered by AT&T; the Bell Labs Museum, administered by Lucent Technologies; and the SHOT History of Computing, administered by the Computer History Association of California. An obituary is in the New York Times (2 Feb. 1995).

Hartley S. Spatt