How Powerful Were the First Supercomputers?

Article Details
  • Written By: Jason C. Chavis
  • Edited By: Bronwyn Harris
  • Images By: Pedro Villavicencio, Sven Gaedtke
  • Last Modified Date: 03 November 2019
  • Copyright Protected:
    Conjecture Corporation
  • Print this Article
Free Widgets for your Site/Blog
In 2009, swimming’s governing body banned the full-body "supersuits" worn by many athletes at the 2008 Olympics.  more...

November 14 ,  1972 :  The Dow Jones closed higher than 1,000 for the first time in history.  more...

The first supercomputers used the basic concept of electronic programming and numerical transactions in order to process information. Using a scalar processing technique, the early designs were very simple by modern standards. Each computer was able to process only one data item at a time, using either the concept of integral data types or floating point numbers. Integral data types within the methodology of computer science represent a finite subset of mathematical integers to form a process. The floating point method is a system in which a string of bits represents a rational number.

Some of the other first supercomputers, however, used vector processors to conduct their data processing. This new central processing unit (CPU) design allowed users to perform operations of multiple mathematical equations simultaneously. Basically, it was able to implement the scalar processing method on a much more elaborate level. This technique was highly acclimated to scientific computing throughout the 1980s, but disappeared with the advent of better CPUs and parallel processing. Parallel processing, utilizing thousands of CPUs, became the basis of modern supercomputers.


Although computers the size of a room had existed since the 1940s, it wasn't until the 1960s that researchers began to fully utilize the concept of design to develop the first supercomputers. The primary scientist to develop the first supercomputers was Seymour Cray at the Control Data Corporation. After building a number of designs, he left the firm in 1980 to establish Cray Research, an independent design company. His designs for supercomputers controlled the market for a number of years before the mass crash of the supercomputer market in the 1990s and his death in 1996.

Considered to be the very first supercomputer, the Cray-1™ system was installed in 1976 at the Los Alamos National Laboratory. At the time, its speed held the world record of 160 million operations per second, also known as megaflops. It contained only 8 megabytes of memory. However, one of the major advances that made the computer durable and powerful was its design, which used wires no more than four feet (1.22 m) in length. In addition, the cooling system used Freon, which prevented overheating.

A short time later, the European Center for Medium-Range Weather Forecasts (ECMWF) obtained the next generation supercomputer, the Cray 1-A™. This was fitted with a system that enabled a 10-day forecast to be processed in five hours. Prior to this supercomputer, the system used by the ECMWF took 12 days to process the same length of forecast. With this installation, the era of the supercomputer as a major tool first took root, changing the world forever.


You might also Like


Discuss this Article

Post your comments

Post Anonymously


forgot password?