Supercomputer site on Lonympics

History of supercomputing

Types of supercomputing and most powerful types of supercomputer

History of supercomputing

The term supercomputer refers to a computer that is leading the world in terms of processing capacity, particularly speed of calculation, when the computer is new. The term Super Computing was used by the New York World newspaper in 1920 to refer to the large custom built tabulators IBM made for Columbia University.

Early history of computing.

Babbage is regarded by many as having made the first computer in 1837. This was a calculation machine. You could say it was half way between a mechanical abacus and a computer. Further developments were made of non digital computing systems using machinery in 19th century. The punch card and the vacuum tube appeared by the end of the 19th century. During the early half of the 20th century, many computing needs were met by sophisticated, special-purpose analog computers, which used direct physical or electrical model of the problem as a basis for computation. Tabulations machines were being construnted by IBM. Computing became more like todays system from the 1930s. In 1937 Claude Shannon invented digital electronics. The secret British Colossus computer demonstrated a device using valves could be made reliable and reprogrammed to do diferent calculations, rather the same types of specialist calculations. The American ENIAC (1946) which was one of the first general purpose machine, but still used the decimal system and incorporated an inflexible architecture that meant reprogramming it essentially required it to be rewired. Then in 1950 Von Neuman developed a computermodel which could predict the weather showing the increasing use of computer systems. Von Neumann (Neumann János) (December 28, 1903 – February 8, 1957) was a Hungarian-German mathematician and polymath of Jewish ancestry who made important contributions in quantum physics, functional analysis, set theory, computer science, economics and many other mathematical fields.

History of supercomputing

Supercomputing started to take off in the 1960s these were mainly developed by Seymour Cray at Control Data Corporation (CDC), and CDC led the market through the 1970s until Cray left to form his own company, Cray Research. Cray started to dominate the supercomputer market with new designs, holding the top spot in supercomputing for 5 years (1985–1990). In the 1980s a large number of smaller competitors entered the market. in a parallel to the creation of the minicomputer market a decade earlier, but many of these disappeared in the mid-1990s "supercomputer market crash". Companiessuch as Kendall Square Research and Thinking Machines which were pionearing companies in the eighties went out of business. Today, supercomputers are typically one of a kind custom designs produced by traditional companies such as IBM and HP, who had purchased many of the 1980s companies to gain their experience, although Cray Inc. still specializes in building supercomputers. Fujitsu, have also been making supercomputing technology. Such as the vector parralel processing technology. Now Blue gene is the dominant supercomputer developed by the United States Department of Energy, industry IBM in particular, and academia.

Types of supercomputing

There are many different ways of classifying a supercomputer. One of interest is vector verus scalar.

A scalar processor processes one data item at a time. In vector processor, by contrast, a single instruction operates simultaneously on multiple data items.

Massively parallel processing supercomputer

In the eighties parallele prcoessing developed where processing is shared between processors to help share the processor work and speed up calculations.

It is a type of parallel computing where hundereds, if not thousands of processing elements are used to work together on an software application in a tightly-coupled fashion. This uses scalar processing. MPP was developed in the eighties by companies such as Kendall Square Research and Thinking Machines, and has become the dominant supercomputer architecture. The fact it is dominant means weather prediction companies wishing to have the highest prcoessing spped have had to reconficute theri prcessing to that which owkrs best on scalr processors. When a vector pcressing systems works best, comopanies may wish toe reconfisdure their pgramming to programming tha works best on vector procressing systems. As a result in weather prediction comopanies using vectorised physics are likely to not as sucessful as those that uise scalar. Since scalar is the dominant processing technique at present.

Distributed computing

This is where different computers are used to share the rocessinf for a task to have the same prcoessing power of a supercomputer. In clsuters. At present culsters are not as powerful as the most powerful supercomputer. Yet a single modern desktop PC is now more powerful than a 15-year old supercomputer

Vector parrallel processing

This uses parallel processing excpet with vector processing. Fujitsu and cray develop thse sort of superompoter techniology.

Quantum computing

Mnay feel that quantum comoputing will be the dominant type of supercomputer in the future, but we will have to wet decades to see.

Supercomputer processing speed is measured in flops. That is floating point operations per second.

Major uses of supercomputer

Weather prediction, climate prediction, millatry operations,NASA Advanced Supercomputing facility use it them for simulating space missions.

Links

http://www.jpbowen.com/publications/thes-cray.html A website looking at the history of Seymour Cray the man behind cray computing

http://en.wikipedia.org/wiki/Supercomputer A site on supercomputing

http://en.wikipedia.org/wiki/BlueGene/L Info on bluegene/l supercomputer

http://www.top500.org/ A website looking at the top 500 supercomputers in the planet.

http://www.top500.org/lists/2005/11/TOP10_Nov2005.pdf The most recent top 10 of supercomputing

http://www.llnl.gov/asci/platforms/bluegenel/ Website on blu gene/l architecuture. at lawrence livermore laboratory.

http://www.proj-mission.org/EE660/Ch8.1-3.doc Scalar processing verus vector processing.

Supercomputer companies

Cluster

Resources, Inc.
Cray Inc.
Fujitsu
Galactic Computing Corp.
Groupe Bull
IBM
nCUBE
NEC Corporation
Supercomputer Systems
SGI

Hewlett-Packard/Intel