Seymour Cray was born on September 28, 1925 in Chippewa Falls, Wisconsin. His farther was a civil engineer and spurred Seymour's interest in science. Seymour was given the basement of his parents' house to use as a laboratory for his studies and research. At the age of ten he designed a device that would generate morse code signals based on punched tape that was passed through it.
In 1943, he graduated from high school and was drafted into the United States Army. His technical skills made him ideal for the job of radio operator and he was sent to Europe. There, he served on the front lines, but survived without injury. Later, he was sent to the Pacific to work on breaking codes that the Japanese were using.
After the war, he returned to the United States and began studying electrical engineering at the University of Minnesota. In 1950 he graduated and found a job at Engineering Research Associates, a firm located in St. Paul, Minnesota. In his spare time he pursued a master's degree in applied mathematics and received it in 1951.
At ERA, Cray began working on digital computers, which were cutting edge at the time. He was one of the leading designers of the ERA 1103, which would become one of the first successful computers for scientific functions. When ERA was acquired by Sperry-Rand Corporation, he was placed in their scientific computing branch.
In 1958, after completing a project for the navy called the Naval Tactical Data System, Cray left to work at Control Data Corporation (CDC). There, he was placed on a team that worked on designing a successor to the ERA 1103 and finished it in 1960, calling it the CDC 1604. Just as the new product started shipping, he set to work on the next version, the CDC 6600.
The 6600 was a hugely powerful machine for its time, largely due to Cray's work on optimizing the entire system. It became known as the first commercial supercomputer and outperformed everything else on the market at the time. In response to weak attempts by competition to outdo his design, Cray designed the CDC 7600, which proved to be 5x faster than the 6600.
By 1970, Cray had become quite tired of interference by upper management and lobbied for a lab to be built in his hometown of Chippewa Falls. His wish was granted and CDC constructed a new lab on land that Cray owned. He built a house for himself nearby that featured a bomb shelter in case of nuclear war.
In the new lab, he started development on the CDC 8600. After he finished the project in 1972, he decided to leave the company and start his own, Cray Research. In 1973, he built a lab on the same property as the CDC lab. Unfortunately, the company did not have nearly enough money to design a new computer, but after some meetings on Wall Street, they were able to secure more than enough funding.
After three years of development, the company's first product was the Cray-1, released in 1976. It became the fastest computer in the world, even beating out CDC's offerings. The first model constructed by the company was leased to Los Alamos in 1976 and generated a lot of interest. Later that year, the first system was sold to the National Center for Atmospheric Research for nearly $9 million. The company managed to sell over one hundred more systems to other companies and Cray Research became a massive success.
Cray set to work on the next models, the Cray-2 and a four-processor model called the Cray X-MP. The X-MP was very popular, but the Cray-2 had mediocre sales since it was not much faster than the X-MP and released later. In 1980, he stepped down from his position as CEO to work in a new lab in Colorado Springs, Colorado on the Cray-3.
After nearly a decade of development, the Cray-3 was having some major issues and the company decided to abandon it and focus on the development of the Cray Y-MP, a successor to the X-MP. Cray decided to spin off his lab from the rest of the company, forming Cray Computer Corporation. He kept the Cray-3 project going, but it proved to be a dismal failure, selling very few models. He kept working on developing the Cray-4, which would run at 1 GHz, but it also failed commercially due to lack of demand. The company was forced to file for bankruptcy in 1995.
That same year, he decided to start a new company, SRC Computers, and began designing a massively parallel supercomputer. However, his work was tragically cut short when he got into a traffic collision on Interstate 25 in Colorado Springs, Colorado. He died on October 5, 1996 from injuries sustained from the accident.