Seymour Cray was born in 1925 in Wisconsin and died in 1996 in Colorado. Seymour Cray was a pioneer in supercomputing. His innovations include vector register technology, cooling technologies, and magnetic amplifiers.
Education:
B.S. Electrical Engineering, University of Minnesota, 1950
M.S. Applied Mathematics, University of Minnesota, 1951
Professional Experience:
Engineering Research Associates, 1950-1957
Control Data Corp., 1957-1972
Cray Research Inc., 1972-1989
Cray Computer Corp., 1989-1995
SRC Computers Inc., 1996
Honors and Awards:
W.W. McDowell Award, American Foundation of Information Processing Societies, 1968
Harry H. Good Memorial Award, 1972
Seymour Cray was born in 1925 in Chippewa Falls, Wisconsin. Cray is an English name. His father was a civil engineer who spent the earlier years of his career working for Northern States Power Company. He surveyed dams before he became the town engineer in Chippewa Falls. His mother provided his father with a social life to balance his technical job. Growing up Cray had always been fascinated with electronics. In high school he would spend much of his time in the electrical engineering laboratories. As a hobby he would toy with electrical equipment of all types: radios, electrical motors, electrical circuits, etc.. Computers were not available for Cray at this point in time. He graduated from high school in 1943 before getting shipped off to war.
Cray got to visit both theaters in the war. He arrived in Europe after the D-Day invasion, but he saw the Battle of the Bulge and battled through Germany to reach the Russians. Next, Cray finished up in the Philippine Islands before returning home and resuming his education. In 1950, Cray received his Bachelor of Science in Electrical Engineering from the University of Minnesota. A year later he received his Masters degree in Applied Mathematics. Taking the advice of a professor, Cray ventured down the street to Engineering Research Associates(ERA).
ERA was one of the first computing facilities in terms of digital circuits. It had been sponsored by the Navy to create cryptographic equipment. When Cray got there, the company was around one year old. One of the most exciting attributes of Cray's new company was that the building which housed the company had previously been a glider factory. The factory was building wooden gliders to be used in the D-Day invasion. So Cray set out to build computers in a wooden glider factory.
Cray started out working on the 1100 series computers which later turned out to be Univac. Since the company was so new, Cray spent a great deal of time researching what computing should be and how it works. He listened to lectures from such scholars as Von Neumann. However, he soon found out that there was not much material on the subject of computers to be researched. So he was led to rely on his own invention.
After being a member of a design group for the development of two computers, Cray was set on his own for the development of the next computer. Cray took to working odd hours to avoid distractions from coworkers. Working odd hours became a permanent part of his life. Cray's theory on designing his computers became one of responding to the criticism of customers. He would address the problems and build on them for the next try. The first computer Cray designed by himself was the 1103 computer. From this generation down the line to his Cray-4, Cray integrated the qualities and the solutions of previous generations of computers.
Cray based his 1103 on simplicity. He did not include anything that was not necessary. Competitors in the business were incorporating bells and whistles into their schemes but Cray stuck to the basics. He used RISC like thinking before RISC was even created. In addition to simplicity, Cray also concentrated on completing the projects quickly. Time was of superior importance than anything else. Of course, speed was still the fundamental concept of building a computer.
After a couple of years, ERA was purchased by Remington Rand. The change of ownership did not change the setting of Cray's work. Things just went on as normal. Later on, Sperry purchased the business, followed by Burroughs, and most recently by Unisys. People and products did not change as much as the ownership changed. However, when Remington took over the focus of products switched to commercially attractive computers. Eventually, Cray and his scientific computers were phased out so Cray decided to join Bill Norris and help start Control Data Corporation.
Seymour Cray was not one of the creators of Control Data Corporation, but he was the first technically oriented worker on the staff. Cray had his goal in mind: he wanted to create fast scientific computers. Despite the other founders thoughts to jump into commercial computers, Cray kept his goal in mind and convinced the others that they would build scientific computers. So Control Data Corporation and Cray began building large scientific computers. Soon they had taken away the market from ERA. Early earnings reached up to 600,000ドル which was a lot of money in that day. The company accomplished this by selling stock on the street corner for a dollar or so. Cray risked 5,000ドル in the company stock.
Cray's dedication to creating scientific computers streamed from his understanding of science. This allowed Cray to focus his goals and direct his efforts towards his work. His exclusion of other fields was a small price to pay for what he achieved for scientific computing. Cray enjoyed improving the currently available tools. He saw his benefit to society, he received positive feedback from coworkers and clients and he enjoyed his work. Cray's work was gratifying to himself.
Seymour Cray had a methodology for creating new computers. First, he would take input from the customers who were using previous models. He would study their complaints and desires. From this information, Cray would figure out how these ideas could be used in his next design. This design process was essentially a solo project for Cray. Of course he needed support technicians but the basic concepts were his and his alone. Cray felt this was the only way to get things done efficiently and correctly. Once completed, this project would be reassessed and a new project would be undertaken. Cray would use his goals to set time and financial budgets. Falling off budget was one of Cray's major issues. Of course, with every project of this magnitude, there exists a risk. Cray was taking risks to ensure that his goals would be reached. Cray's goal was to create the fastest scientific computer ever. Cray would design the basics of the computers. He would address such issues as, "What are the instructions in the computer?", "How big is the memory?", and "What is the memory made of?". Once this was decided, the support technicians could create the computer.
As CDC became a bigger company and gained more business, Cray found the distractions to be overwhelming. So he decided to get out of town. He moved to Chippewa Falls and had a research and development facility built there Cray's first project at CDC was the1604 Computer. Cray had decided to create this computer using transistors which were still fairly new. Of course the transistors are expensive, so Cray gets reject transistors from a local retail store for cheap. Cray proved that substandard components could be designed around so that the goal of a working computer can be achieved. The 1604 was very successful in the market because it was the only large computer using transistors at the time. Next for Cray came the 6600 and the 7600 which were the most revolutionary projects for Cray to date, while still maintaining the scientific focus. Scientists were discovering how to solve partial differential equations in an iterative process, so more iterations meant more precise answers. This means faster computers were required. So the 6600 and 7600 came of age.
In 1972, Cray realized that the market for large scale scientific computers had become small so soon CDC would be discontinuing their efforts in this area. So Cray decided that 1972 would be a good time to open up a small business to take over this market. Thus was the beginning of Cray Research Inc.. Cray used the newest technology, integrated circuits and vector register technology, to build Cray 1. In designing the Cray 1, Cray was able to express himself artistically by giving the Cray 1 an attractive appearance. It was conical in shape so the wires would be on the short inner surfaces of the modules. Cray went on to design the next immensely successful Cray Research computer, the Cray 2. The Cray 3 computer could reach speeds up to 500MHz but none were ever sold. The Cray 4 was twice as fast as the previous but was never completed. After this successful business had given Cray his fill, he went on to create one more company. He called it Cray Computer Corporation. This business failed, so about a year and a half later in 1996, Cray tried one last time to build a company which would build super fast scientific computers. But...
On September 22, 1996 around 3:00PM, Seymour Cray was merging onto southbound I-25 from North Academy Boulevard in Colorado. A Chevrolet Camaro was following behind him. Both cars merged into the left lane, but the Camaro tried to switch into the left lane and almost hit a Pontiac Grand Prix in the left lane. The Grand Prix swerved off the road onto the shoulder then back onto the road and hit the Camaro. The Camaro spun onto the medium, but the Grand Prix swerved more and rammed the left-rear door of Cray's Cherokee. The Cherokee spun counter-clockwise before rolling three times and coming to a halt in the middle of the lane. Cray suffered a broken neck, broken ribs, and severe head injuries. Cray was taken in an ambulance to Penrose Hospital where he received surgery to relieve brain swelling. Cray remained in critical and unstable condition until October 5th when Seymour Cray was pronounced dead at 2:53AM. By his thirties, Cray had been established as a genius in designing fast scientific computers. Seymour Cray's name will forever be synonymous with high performance computing.
Quotations:
"He is the Thomas Edison of the supercomputing industry," --Larry L. Smarr
When told that Steve Jobs bought a CRAY to help design the next Apple, Seymour Cray said, "Funny, I am using an Apple to simulate the CRAY-3."
Bibliography
Biographical Publications
Interview with Seymour Cray http://innovate.si.edu/history/cray/craytoc.htm
GT OnLine Top Stories http://www.usa.net/gtonline/archive/96-09-23/top010.html
Seymour Stories by thread http://www.cray.com/PUBLIC/SEYMOUR/STORIES/