Every year in Linux Adictos We echo the publication of the list of the 500 largest supercomputers in the world. And in a previous article I told you about Frontier, which at the moment is not only the fastest in the world but also the best performing in the world.
How to Live Aligned with it's always interesting to know how we got here, let's go with a brief history of supercomputers.
What is a supercomputer?
It is a team capable of executing long and complex calculations at enormous speed. The performance of a supercomputer is commonly measured in floating point operations per second (FLOPS) rather than millions of instructions per second (MIPS).
Supercomputers are used in a wide field of activities that require intensive use of computation. such as quantum mechanics, weather forecasting, climate research, oil and gas exploration, molecular modeling (computing the structures and properties of chemical compounds, biological macromolecules, polymers, and crystals), and computer-type theoretical physics simulations. simulations of the first moments of the universe, aerodynamics of aircraft and spacecraft, detonation of nuclear weapons and nuclear fusion. Also in the creation and breaking of secure encryption methods.
Brief history of supercomputers
In 1956, a team at the University of Manchester in the UK began developing MUSE. His goal was to build a computer that could operate at processing speeds close to one microsecond per instruction, that is, about a million instructions per second. Shortly thereafter, the name of the project was changed to Atlas.
The first Atlas was officially commissioned on December 7, 1962, and was considered at the time of its launch to be the most powerful computer in the world. The Atlas pioneered the use of virtual memory and paging as a way to expand its working memory by combining its 16 words of main memory and an additional 384 words. secondary battery memory.
The first computer that came from a private company was from a company founded in 1957 by a group of computer engineers, including Seymour Cray, who would later become one of the most prominent personalities in the industry. The company was called Control Data Corporation and produced a computer called CDC 6000 that was made up of four hundred thousand transistors, one hundred miles of wiring, an innovative cooling system and a record computing power for the time of 3 megaFLOPS. This computer appeared in 1964
The secret to the speed of the CDC 6600 was in its ability to share work with peripherals allowing the CPU to be dedicated solely to data processing. The programming language was FORTRAN.
In 1968, Cray produced the CDC 7600 which also achieved the title of the fastest computer in the world.. Running at 36 MHz, the 7600 had 3,6 times the clock speed of the 6600, but was not as commercially successful as expected. and Cray left CDC in 1972 to form his own company.
Without your involvement, CDC produced the STAR-100 with a speed of 100 megaFLOPS, that is three times the speed of its predecessor. and used the so-called vector processing, that is, the CPU was designed to perform multiple mathematical calculations simultaneously.
Already in his own company, Seymour Cray produced three models
- CRAY-1: It dates from 1976 and was one of the first supercomputers to use integrated circuits and worked at a speed of 160 megaFLOPS.
- CRAY X-MP: It appeared in 1982 adding 4 processors and more memory bandwidth to the previous model. Its calculation capacity is 800 megaFLOPS.
- CRAY-2: This 1985 computer had liquid cooling and a calculation speed of 1,9 gigaFLOPS.
Like many other pioneers, Cray failed to detect the paradigm shift and his company filed for bankruptcy in 1995. Meanwhile, its competitors embraced the current model of parallel computing in which a task is divided between two or more processors that are responsible for solving it simultaneously.