Skip to main content
Select Source:

Seymour Cray

Seymour Cray

Seymour Cray (1925-1996) is one of the founding fathers of the computer industry. Seeking to process vast amounts of mathematical data needed to simulate physical phenomena, Cray built what many consider the first supercomputer, which represented a technological revolution to such fields as engineering, meteorology, and eventually biology and medicine.

Seymour Cray is an electronics engineer and one of the founding fathers of the computer industry. His seminal work in computer design features the semiconductor as a component to store and process information. Cray's dense packing of hundreds of thousands of semiconductor chips, which reduced the distance between signals, enabled him to pioneer very large and powerful "supercomputers." Among his accomplishments was the first computer to employ a freon cooling system to prevent chips from overheating. However, Cray's most significant contribution was the supercomputer itself. Seeking to process vast amounts of mathematical data needed to simulate physical phenomena, Cray built what many consider the first supercomputer, the CDC 6600 (with 350,000 transistors). To such fields as engineering, meteorology, and eventually biology and medicine, the supercomputer represented a technological revolution, akin to replacing a wagon with a sports car in terms of accelerating research.

A maverick in both his scientific and business pursuits, Cray eventually started his own company devoted entirely to the development of supercomputers. For many years Cray computers dominated the supercomputer industry. A devoted fan of "Star Trek," a 1960s television show about space travel, Cray included aesthetically pleasing touches in his computers, such as transparent blue glass that revealed their inner workings.

Cray was born on September 28, 1925, in Chippewa Falls, Wisconsin, a small town situated in the heart of Wisconsin's dairy farm country. The eldest of two children, Cray revealed his talent for engineering while still a young boy, tinkering with radios in the basement and building an automatic telegraph machine by the time he was ten years old. Cray's father, a city engineer, and his mother fully supported his scientific interests, providing him with a basement laboratory equipped with chemistry sets and radio gear. Cray's early aptitude for electronics was evident when he wired his laboratory to his bedroom, and included an electric alarm that sounded whenever anyone tried to enter his inner sanctum. While attending Chippewa Falls High School, Cray sometimes taught the physics class in his teacher's absence. During his senior year, he received the Bausch & Lomb Science Award for meritorious achievement in science.

While serving in the U.S. Army during the final years of World War II, Cray utilized his natural gifts in electronics as a radio operator and decipherer of enemy codes. After the war, he enrolled in the University of Wisconsin, but later transferred to the University of Minnesota in Minneapolis, where he received his bachelor's degree in electrical engineering in 1950 and a master's degree in applied mathematics the next year. Cray began his corporate electronics career when he was hired to work for Engineering Research Associates (ERA). When Cray joined the company, it was among a small group of firms on the cutting edge of the commercial computer industry. One of his first assignments with ERA was to build computer pulse transformers for Navy use. Cray credited his success on the project to a top-of-the-line circular slide rule that enabled him to make a multitude of calculations needed to build the transformers. In a speech before his colleagues at a 1988 supercomputer conference, Cray recalled feeling "quite smug" about his accomplishment until he encountered a more experienced engineer working at the firm who told Cray that he did not use complicated slide rules or many of the other standard engineering approaches in his work, preferring to rely on intuition. Intrigued, Cray put away his slide rule and decided that he would do likewise.

For his next computer project, Cray and his colleagues developed a binary programming system. With the addition of magnetic core memory, which allowed Cray and his coworkers to program 4,096 words, the age of the supercomputer dawned. Although devoted to his laboratory work, Cray was also interested in the business side of the industry; his efforts to market ERA's new technology resulted in the Remington Rand typewriter company buying out ERA. With a formidable knowledge of circuits, logic, and computer software design, Cray designed the UNIVAC 1103, the first electronically digital computer to become commercially available.

Despite his growing success, Cray became dissatisfied with the large corporate atmosphere of ERA, which had been renamed the Sperry Rand Corporation. A friend and colleague, William Norris, who also worked at Sperry Rand, decided to start his own company, Control Data Corporation (CDC), and recruited Cray to work for him. Lacking the financial resources of larger companies, Cray and Control Data set out to make affordable computers. Towards this end, Cray built computers out of transistors, which he purchased at an electronics outlet store for 37 cents each. Although the chips were of diverse circuitry, Cray successfully replaced the cumbersome and expensive tubes and radio "valves" which were then standard in the industry.

Control Data began developing a line of computers like the CDC 1604, which was immensely successful as a tool for scientific research. Cray went on to develop the CDC 6600, the most powerful computer of its day and the first to employ freon to cool its 350,000 transistors. In 1969, the corporation introduced the CDC 7600, which many considered to be the world's first supercomputer. Capable of 15 million computations per second, the 7600 placed CDC as the leader in the supercomputer industry to the chagrin of the IBM corporation, CDC's primary competitor. Even with a legion of researchers, IBM was unable to match CDC's productivity, and eventually resorted to questionable tactics to overtake CDC, which eventually filed and won an antitrust suit against IBM. But as Control Data grew, so did its bureaucracy. As Russell Mitchell recounted in Business Week, Norris once asked Cray to develop a five-year plan. What Norris received in return was a short note that said Cray's five-year plan was "to build the biggest computer in the world," and his one-year plan was "to achieve one-fifth of the above." After developing the CDC 8600, which the company refused to market, Cray, in 1972, decided to leave CDC and set up his own company, Cray Research Corporation. Norris and CDC graciously invested $500,000 to assist Cray in his fledgling business effort.

Cray Research immediately set out to build the fastest supercomputer. In 1976 the CRAY-1 was introduced. Incorporating a revolutionary vector processing approach, which allowed the computer to solve various parts of a problem at once, the CRAY-1 was capable of performing 32 calculations simultaneously, outpacing even the best CDC computer. When the National Center for Atmospheric Research met the computer's $8.8 million price tag, Cray Research finally had solid financial footing to continue building faster and more affordable computers. For Cray, this meant manufacturing one product at a time, a radical approach in the computer industry. The first CRAY-2 was marketed in 1985 and featured a phenomenal 2-billion byte memory that could perform 1.2 billion computer operations per second, a tenfold performance increase over the Cray-1. Capable of providing computerized models of physical phenomena described mathematically, the CRAY computers were essential catalysts in accelerating research. For example, in such areas as pharmaceutical development, supercomputer modeling of a drug's molecules and its biological components eliminated much trial and error, reducing the time necessary to solve complicated mathematical equations.

In 1983, Cray turned his attention to developing gallium arsenide (GaA) circuits. Although the CRAY-2 was based on silicon chips, Cray continued to develop GaA chips in the spinoff Cray Computers Corporation. Although extremely difficult to work with because of their fragility, gallium arsenide computer chips marked a major advance in computer circuitry with their ability to conduct electrical impulses with less resistance than silicon. Adding even more speed to the computer, the GaA chip also effectively reduced both heat and energy loss.

While Cray's advances in computer technology enabled him to corner the market on the supercomputer industry for many years, the advent of parallel processing allowed others in the industry to make inroads into the same market. Utilizing hundreds of mini-computers to work on individual aspects of a problem, parallel processing is a less expensive approach to solving huge mathematical problems. Although Cray for many years denounced parallel processing as impractical, he eventually accepted this approach and made plans with other companies to incorporate it into his computer research and business.

Cray's first wife, Verene, was a minister's daughter. Married shortly after World War II, they had two daughters and two sons, who have characterized their father as a man intensely dedicated to his work; in fact, Cray demanded their absolute silence while traveling in the car so that he could think about the next advance in supercomputers. In 1975, Cray and Verene divorced, and he wed Geri M. Harrand five years later. Although he engaged in outdoor pursuits with his new wife, such as windsurfing and skiing, Cray remained devoted to his research. In 1972, he was awarded the Harry Goode Memorial Award for "outstanding achievement in the field of information processing." As Cray looked forward to the future of supercomputers, especially to the use of GaA computer chips, many experts in the field characterized his vision as impractical. Nonetheless, Cray's numerous conceptual breakthroughs in computer and information science have firmly established him as an innovator in computer technology. Cray died on October 5, 1996, from injuries sustained in a car accident three weeks earlier.

Further Reading

Slater, R, Portraits in Silicon, MIT Press, 1989, pp. 195-204.

Spenser, Donald, Macmillan Encyclopedia of Computers, Macmillan Publishing Company, 1992.

Anthes, Gary H, "Seymour Cray: Reclusive Genius," in Computerworld, June 22, 1992, p. 38.

Elmer-Dewitt, Philip, "Computer Chip Off the Old Block: Genius Seymour Cray and the Company He Founded Split Up," in Time, May 29, 1989, p. 70.

Krepchin, Ira, "Da tamation 100 North American Profiles," in Datamation, June 15, 1993, p. 81.

Mitchell, Russell, "The Genius," in Business Week, April 30, 1990, pp. 80-88. □

Cite this article
Pick a style below, and copy the text for your bibliography.

  • MLA
  • Chicago
  • APA

"Seymour Cray." Encyclopedia of World Biography. . Encyclopedia.com. 22 Aug. 2017 <http://www.encyclopedia.com>.

"Seymour Cray." Encyclopedia of World Biography. . Encyclopedia.com. (August 22, 2017). http://www.encyclopedia.com/history/encyclopedias-almanacs-transcripts-and-maps/seymour-cray

"Seymour Cray." Encyclopedia of World Biography. . Retrieved August 22, 2017 from Encyclopedia.com: http://www.encyclopedia.com/history/encyclopedias-almanacs-transcripts-and-maps/seymour-cray

Cray, Seymour

Cray, Seymour

American Computer Designer
19251996

Seymour Cray, widely regarded as the father of supercomputers , was born in Chippewa Falls, Wisconsin, on September 28, 1925. He died on October 5, 1996, after a car accident. A man of unerring focus, Cray maintained a single goal throughout his lifetimeto build the fastest computer possible.

After obtaining a bachelor's degree in electrical engineering and a master's degree in applied mathematics from the University of Minnesota, Cray started his career with Engineering Research Associates, which became Remington Rand, then Sperry Rand. Here Cray designed his first computer, the ERA 1101. He also helped design the UNIVAC (Universal Automatic Computer) 1103. When the UNIVAC shifted its focus away from supercomputers, Cray did not. Instead, he left to start Control Data Corporation (CDC), leading the first design team to abandon vacuum tubes and produce the first commercial transistor -based computer, the CDC 1604, in 1958.

Within a few years, Cray decided that the administrative duties associated with running a successful business interfered with his concentration on computer design. With CDC's blessing, he left for home and solitude in Chippewa Falls. CDC built Cray's laboratory close to his home. His time there proved productive, and in 1963, CDC released the Cray-designed CDC 6600. To reduce heat build-up in the compact unit, Cray installed a freon cooling system inside the computer. This machine was not only three times faster than its IBM counterpart, but was also smaller and a bargain at $7.5 million each. Customers for the 6600 included the U.S. Atomic Energy Commission, which used it to to study thermonuclear explosions, and the U.S. Weather Bureau, which used it to to assist in weather predictions.

In 1969 CDC released the Cray-designed 7600, which could perform twenty megaflops, or twenty million calculations (floating point operations or FLOPS) per second. Many believe the 7600 was the first supercomputer.

When CDC declined to market Cray's next machine (the 8600), instead expanding into other commercial ventures, Cray again chose to focus on supercomputer design. He started Cray Research in 1972. The Cray 1, released in 1976, was the first computer to master successfully vector processing and at the time was the fastest scalar processor in the world. This single-processor, 160 megaflop computer had 1 million, 64-bit words of central memory, and used three types of integrated circuits . The computer itself was an 8 6 foot (about 2.4 1.8 meter) cylindrical structure. With a maximum wire length of 4 feet (1.2 meters), it still required a total of 60 miles (18.3 meters) of wire. Each unit cost $8.8 million. Cray Research went public in 1976, a month before its first sale (to Los Alamos National Laboratories in New Mexico). The sale of stock raised $10 million, helping the company to survive.

By 1984 Cray Research owned 70 percent of the supercomputer market. Cray himself was worth $17 million. In November of 1981, Cray resigned as CEO of Cray Research to become an independent contractor. Again, he sought solitude to build what he called "simple, dumb things." While he worked on the Cray 2, Cray Research, led by Steve Chen, revised the Cray 1. Chen's resultant Cray X-MP was the first multi-processing computer, able to perform 500 megaflops. The resultant rivalry between Cray and Chen eventually led to Cray's departure from the company in 1989. Cray then founded Cray Computer Corporation of Colorado Springs.

Cray wanted the Cray 2 to be four to six times more powerful than the Cray 1. When finished in 1985 with four processors, the Cray 2 doubled that expectation. This supercomputer had the largest internal memory capacity of any computer (two million bytes) and 256 million words of central memory. It could perform 1.2 gigaflops (or 1.2 million FLOPS per second). Cray fit the entire structure in a C-shaped cabinet 53 inches (136 centimeters) across and 45 inches (115 centimeters) high. To diffuse heat, he immersed the circuits in the coolant Fluorinert. Its cost was $17.6 million. What the Cray 2 accomplished in one second had taken a year to accomplish by hand in 1952. This computer was used to study intense magnetic fields associated with fusion reactors and to help design heat shields for space exploration.

In 1985 Cray began work on the Cray 3, using 16 processors. This machine was based on gallium arsenide instead of silicon , but the design was eventually dropped due to its high development cost. He then began work on the Cray 4, which would have had 64 processors. This project, too, was dropped. Cray Computer filed for bankruptcy protection in 1995.

Still, Cray was convinced that he could design a high performance, low cost system. By the time of his sudden death in 1996, he had established another new company, SRC Computers, toward this goal.

Two factors are cited in the decline of the supercomputer market. First, competition with cheaper and increasingly powerful microprocessors, and the ability for minicomputers to be networked, helped reduce the demand for very expensive machines. Second, with the end of the "Cold War," military budgets were slashed. New supercomputers for military use were no longer a practical consideration, reducing the already exclusive market.

see also Minicomputers; Physics; Supercomputers.

Mary McIver Puthawala

Bibliography

Markoff, John. "Seymour Cray, Computer Industry Pioneer and Father of the Supercomputer, Dies at 71." New York Times, October, 6, 1996, p. 1.

Slater, Robert. "Seymour Cray: The Hermit of Chippewa Falls and His Simple, 'Dumb Things.'" In Portraits in Silicon. Cambridge, MA: MIT Press, 1987.

Internet Resources

Breckenridge, Charles W. "A Tribute to Seymour Cray, SRC Computers, Inc." <http://cgl.ucsf.edu/home/tef/cray/tribute.html> A History of Cray. Cray Inc. <http://www.cray.com/company/> "

Power Mac G4 Cube." Apple Computer, Inc. <http://www.apple.com/powermaccube/>

Cite this article
Pick a style below, and copy the text for your bibliography.

  • MLA
  • Chicago
  • APA

"Cray, Seymour." Computer Sciences. . Encyclopedia.com. 22 Aug. 2017 <http://www.encyclopedia.com>.

"Cray, Seymour." Computer Sciences. . Encyclopedia.com. (August 22, 2017). http://www.encyclopedia.com/computing/news-wires-white-papers-and-books/cray-seymour

"Cray, Seymour." Computer Sciences. . Retrieved August 22, 2017 from Encyclopedia.com: http://www.encyclopedia.com/computing/news-wires-white-papers-and-books/cray-seymour