Howard Aiken

views updated

Howard Aiken

A noted physicist and Harvard professor, Howard Aiken (1900-1973) designed and built the Mark I calculator in the late 1930s and early 1940s. The first large-scale digital calculator, the Mark I provided the impetus for larger and more advanced computing machines. Aiken's later conceptions, the Mark II, Mark III, and Mark IV, each surpassed its previous model in terms of speed and calculating capacity.

Howard Hathaway Aiken was born on March 8, 1900, in Hoboken, New Jersey, and was raised in Indianapolis, Indiana. Because of his family's limited resources, he had to go to work after completing the eighth grade. He worked twelve-hour shifts at night, seven days a week, as a switchboard operator for the Indianapolis Light and Heat Company. During the day he attended Arsenal Technical High School. When the school superintendent learned of his round-the-clock work and study schedule, he arranged a series of special tests that enabled Aiken to graduate early. In 1919, Aiken entered the University of Wisconsin at Madison and worked part-time for Madison's gas company while he attended classes. He received his Bachelor of Science degree in 1923 and, upon graduation, was immediately promoted to chief engineer at Madison's gas company. Over the next twelve years he became a professor at the University of Miami and later went into business for himself. By 1935, he decided to return to school. Aiken began his graduate studies at the University of Chicago before going on to Harvard. He received a master's degree in physics in 1937 and was made an instructor. He wrote his dissertation while he was teaching and received his doctorate in 1939.

Proposed Design for First Modern

Computer As a graduate student in physics, Aiken completed a great deal of work, requiring many hours of long and tedious calculations. It was at that time that he began to think seriously about improving calculating machines to reduce the time needed for figuring large numerical sequences. In 1937, while at Harvard, Aiken wrote a 22-page memorandum proposing the initial design for a computer. His idea was to build a computer from existing hardware with electromagnetic components controlled by coded sequences of instructions—and one that would operate automatically after a particular process had been developed. Aiken proposed that the punched-card calculators then in use (which could carry out only one arithmetic operation at a time) could be modified to become fully automated and to carry out a wide range of arithmetic and mathematical functions. His original design was inspired by the description of a more powerful calculator in the work of Charles Babbage, an English mathematician who had devoted nearly forty years to developing a calculating machine.

Although Aiken was by then an instructor at Harvard (and was to become an associate professor of applied mathematics in 1941 and a full professor in 1946), the university offered little support for his initial idea. He therefore turned to private industry for assistance. Although his first attempt to muster corporate support was turned down by the Monroe Calculating Machine Company, its chief engineer, G. C. Chase, approved of Aiken's proposal. He suggested that Aiken contact Theodore Brown, a professor at the Harvard Business School. Brown, in turn, put Aiken in touch with someone at IBM. Aiken's idea impressed the IBM executive enough that he agreed to finance the construction of what became known as the Mark I. In 1939, IBM president, Thomas Watson, Sr., agreed to build the computer under Aiken's supervision, with additional financial backing from the U.S. Navy. At the time IBM only manufactured office machines, but its management wanted to encourage research in new and promising areas and was eager to establish a connection with Harvard. During that same year, Aiken became a officer of the Naval Warfare School at Yorktown. When the Mark I contract was worked out he was made officer in charge of the U.S. Navy Computing Project. The Navy agreed to support Aiken's computer because the Mark I offered a great deal of potential for expediting the complex mathematical calculations involved in aiming long-range guns onboard ship. The Mark I provided a solution to the problem by calculating gun trajectories in a matter of minutes.

Built Mark I-IV

With a grant from IBM and a Navy contract, Aiken and a team headed by Clair D. Lake began work at IBM's laboratories in Endicott, New York. Aiken's machine was electromechanical—mechanical parts, electrically controlled—and used ordinary telephone relays that enabled electrical currents to be switched on or off. The computer consisted of thousands of relays and other components, all assembled in a 51-foot-long and 8-foot-high (1554 cm x 243 cm) stainless steel and glass frame that was completed in 1943 and installed at Harvard a year later. Seventy-two rotating registers formed the heart of this huge machine, each of which could store a positive or negative 23-digit number. The telephone relays established communication between the registers. Instructions and data input were entered into the computer by means of continuous strips of IBM punch-card paper. Two electrical typewriters hooked up to the machine printed output. The Mark I did not resemble modern computers, either in appearance or in principles of operation. The machine had no keyboard, for instance, but was operated with approximately 1,400 rotary switches that had to be adjusted to set up a run. Seemingly clumsy by today's computer standards, the Mark I nevertheless was a powerful improvement over its predecessors in terms of the speed at which it performed a host of complex mathematical calculations. Many scientists and engineers were eager for time on the machine, underscoring the project's success and giving added impetus for continued work on improved models. However, a dispute developed with IBM over credit for the computer. Subsequently, the company withdrew support for all further efforts. A more powerful model was soon undertaken under pressure from competition from ENIAC, the much faster computer then being built at Columbia University.

Mark I was to have three successors, Mark II through IV. It was with the Mark III that Aiken began building electronic machines. He had a conservative outlook with respect to electronic engineering and sacrificed the speed associated with electronic technology for the dependability of mechanics. Only after World War II did he begin to feel comfortable using electronic hardware. In 1949, Aiken finished the Mark III with the incorporation of electronic components. Data and instructions were stored on magnetic drums with a capacity of 4,350 sixteen-bit words and roughly 4,000 instructions. With Aiken's continued concern for reliability over speed, he called his Mark III "the slowest all-electronic machine in the world," as quoted by David Ritchie in The Computer Pioneers: The Making of the Modern Computer. The Mark III's final version, however, was not completely electronic; it still contained about 2,000 mechanical relays in addition to its electronic components. The Mark IV, which followed on the heels of the Mark III, was considerably faster.

Aiken contributed to the early computing years by demonstrating that a large, calculating computer could not only be built but could also provide the scientific world with high-powered, speedy mathematical solutions to a plethora of problems. Aiken remained at Harvard until 1961, when he moved to Fort Lauderdale, Florida. He went on to help establish a computer science program and computing center at the University of Miami, where he became Distinguished Professor of Information. At the same time he founded a New York-based consulting firm, Howard Aiken Industries Incorporated. Aiken disliked the idea of patents and was known for sharing his work with others. He died on March 14, 1973.