History In

History In the years 1968 to 1972 Cray at Control Data was working on a new machine known as the CDC 8600, the logical successor to his earlier CDC 6600 and CDC 7600. The 8600 was essentially composed of four 7600s in a box. Jim Thornton, former Cray engineering partner in previous designs, had begun a radical project known as the CDC STAR-100. Unlike 8600, he used brute force, STAR took a totally different route. In fact, the main processor of the STAR had less performance than 7600, but added additional hardware and instructions to speed up work, something common in supercomputers. In 1972, the 8600 had reached an impasse. The machine was so incredibly complex that it was impossible to make it work well, even a single faulty component would make the machine operational. Cray went to William Norris, CEO of Control Data, saying it was necessary to redesign the machine from scratch.At a time when the company was in serious financial trouble, and with the STAR on the assembly line, Norris simply could not invest money. Cray left Control Data and starting a new business a few yards from the headquarters of the CDC lab in the backyard of the property he bought in Chippewa Falls, WI, Cray and a group of former employees of the CDC began looking ideas. At first, the concept of building another supercomputer seemed impossible, but after the Technical Director traveled to Wall Street and found a group of investors more than willing to support Cray, all that was needed was a design. In 1975, the Cray-1 80 MHz was announced. The expectation was so high that a war started by the first machine between the Lawrence Livermore National Laboratory and Los Alamos National Laboratory, the latter being the winner and received the machine serial number 001 1976 for a trial period of six months .The National Center for Atmospheric Research (NCAR) was Cray Research first official client in July 1977, paying U.S. 8.86 million ( 7.9 million plus 1 million for the disks). NCAR machine was dropped in January 1979. The company hoped to sell a dozen machines, but they sold more than 80 Cray-1 in all models, with prices ranging from 5M to 8M. The Cray machine turned into a celebrity, and successful company until the fall of supercomputers in early 1990. The Cray-1 was succeeded in 1982 by the Cray X-MP of 800 MFLOPS, the first multi-processing computer. In 1985, the most advanced Cray-2, capable of reaching peaks of 1.9 GFLOPS, succeeded the first two models, but succeeded only limited commercial success because of some problems in achieving a sustained performance in the world applications real. Therefore, there was a more conservative design approach as the successor of the Cray-1 and X-MP, called the Cray Y-MP, released in 1988.