Whatever Happened to the Seven Dwarfs.
Dwarf Three: Control Data
by John C. Dvorak
In 1957 the politics at Univac were too much for William C. Morris and he decided he could do a better job of running a computer company than they could at Sperry-Rand, makers of Univac. So Control Data was formed and rolled out its first machine in April of 1958. Sperry sued immediately and settled out of court 4 years later. The first machine was the famous CDC 1604. It was designed by the legendary Seymour Cray and the story of Control Data is essentially the story of Cray.
Seymour Cray, after WWII went to work for ERA — Engineering Research Associates — a company that may have invented many of the first patentable computer designs. The company was founded by a group of Naval intelligence scientists whose goal was to commercialize the technologies that were used to crack Nazi cryptography. Meanwhile, the company that argued that it’s people (Mauchley and Eckert) invented the electronic digital computer itself, Remington-Rand, bought out ERA in 1952 so there would be no dispute. This was akin to Lotus buying out VisiCorp. With the purchase of ERA came Seymour Cray who was easily recognized as a genius by Norris and the band of men who left Remington Rand to form Control Data. Cray’s contribution to UNIVAC was the invention of the Transtec Computer, a fully transistorized machine that was never marketed and some work on the 1103. Cray met John von Neumann at ERA and there was already bickering between von Neumann and Mauchley over various inventions. Eckert and Norris were bickering too. The likelihood that Cray, who liked solitude and calm, would work into the corporate culture at Sperry-Rand was nil. Soon after Control Data was formed Cray joined in.
Cray was one of the first people to recognize the importance of the transistor to computer design and the first machine from CDC was the all-transistor 1604 — the most powerful computer of its era. Instantly in the world of high-performance computing CDC was a player. This machine could do just about everything from real-time data processing to weapons control. It is one of the first machines to be used for teaching the new curriculum of computer science. The 1604 had a clock speed of 5 microseconds. Cray was 35 at the time.
The next Cray machine from CDC is considered by some to jokingly be the world’s first personal computer — the 160A. Using the same technologies as the 1604 it was a dedicated machine used for applications such as production control and various scientific applications. It cost $500,000 and Cray once told an audience that he had designed it from scratch in one week.
In the Cray Museum which houses a 160A there is a framed dollar bill. In the 1950′s and 1960′s it was traditional for a company to own the patents developed in-house by employees. As a reward for a patent the inventor would get a crisp dollar bill framed. This bill was one of the many given to Cray for his work at CDC.
From there Cray led the design of numerous CDC machine right up through the famous 6000 series, which some people consider the first supercomputer. It ran at a then astonishing speed of 3 mips, Not bad for 1963. It was so much faster than the IBM 7094 that apparently Thomas Watson got bent out of shape at IBM over the fact that the 6600 design team was so small and IBM’s was so large.
This is in large part to a system of design later dubbed the Cray Way whereby you’d establish a small highly networked group with a diverse background and have them form a small cooperative team. Cray also isolated his research group away from the corporate headquarters.
The 6600 was revolutionary. It could execute 10 instructions simultaneously, used peripheral processors and multiple data channels. It was the first freon cooled computer. Later in life. as Cray employed more and more liquid cooling techniques to his computer designs, he described the process of computer engineering as “mostly plumbing.”
After the 6600 came the 7600 which was the first machine actually dubbed a supercomputer when introduced. When CDC rejected his 8600 machine in favor of the CDC CYBER line of machines Cray and some of the founders of CDC (who were all with ERA then Sperry originally) left in 1973 to form Cray Research where Seymour Cray designed the Cray-1 and subsequent machines all of which surpassed CDC machines in performance.
As CDC got big it turned into the same kind of bureaucracy that Sperry-Rand had been and which drove the CDC founders to go elsewhere in the first place. The decision to let Cray walk had to be politically involved. Cray apparently wanted nothing to do with any corporate crap. A famous story about him revolves around his being asked to produce a five-year plan for CDC. His report was simple: “Five-year goal: Build the biggest computer in the world. One-year goal: Achieve one-fifth of the above.”
Control Data Became the largest computer maker in Minnesota. The company maintained its line of CYBER machines with the 960 being introduced in 1988. The company went on to design high performance disk drives and other devices but never achieved stardom with any of its computers after it lost Cray. Over time the company toyed with the PLATO teaching machines, got into real estate, and brought out an obscure microcomputer, the CDC 110 in 1981. In 1986 William Norris retired. Sales declined and in 1992 the company split into the Ceridian Corporation and Control Data Systems, Inc.
What was left of Control Data was bought last September by the investment firm of Welsh, Carson, Anderson & Stowe. The new owners say that they are “committed to nurturing Control Data’s focus on providing electronic commerce solutions to large organizations worldwide.” Hardly exciting. What was once the world’s most charismatic maker of supercomputers is now just another E-commerce company.