Supercomputer

From Wikipedia, the free encyclopedia
Jump to: navigation, search
The Cray-2, the world's fastest supercomputer from 1985 to 1989

A supercomputer is a computer with great speed and memory. This kind of computer can do jobs faster than any other computer of its generation. They are usually thousands of times faster than ordinary personal computers made at that time. Supercomputers can do calculation jobs very fast, so they are used for weather forecasting, code-breaking, genetic analysis and other jobs that need many calculations. When new computers of all classes become more powerful, new ordinary computers are made with powers that only supercomputers had in the past, while new supercomputers continue to outclass them.

Electrical engineers make supercomputers that link many thousands of microprocessors.

How it works[change | edit source]

Supercomputer types are: shared memory, distributed memory and array. Supercomputer that has shared memory is developed by using a parallel handling and pipelining concept. Supercomputer that has distributed memory consist of many(about 100~10000) nodes. CRAY series of CRAYRESERCH and VP 2400/40, NEC의 SX-3 of HUCIS is shared memory type. nCube 3, iPSC/860, AP 1000, NCR 3700, Paragon XP/S, CM-5 are distributed memory type.

History[change | edit source]

Array type computer ILIAC started at 1972 and since that, CF-11, CM-2, Mas Par MP-2(that has array type) is developed. Lately, array type supercomputer is beginning to make a mark. And movements that use physically separated memory like one shared memory bring T3D, KSR1, and Tera Computer.

Smallest supercomputer is created[change | edit source]

China's first personal high-performance computer (PHPC100) makes debut in Tianjin. It is the China's smallest supercomputer, only twice the size of a normal desktop computer, but its computing speed is 40 times faster than a normal computer.

The invention of the PHPC100, by the Dawning Information Industry Co. at Tianjin Base, signifies that supercomputers, previously a privilege in high-end research, are no longer exclusive to major fields such as defense-related scientific research.

It perfectly suits the scientific computing, data analysis, graphic simulation and other powerful computing needs of university laboratories or scientific research institutions.

At the same time, it solves problems common in traditional high-performance computer clusters, including power consumption, space occupancy and heat dissipation.

Supercomputing centers, organizations[change | edit source]

Organizations

  • DEISA Distributed European Infrastructure for Supercomputing Applications, a facility integrating eleven European supercomputing centers.
  • NAREGI Japanese NAtional REsearch Grid Initiative involving several supercomputer centers
  • TeraGrid, a national facility integrating nine US supercomputing centers

Centers

Specific machines, general-purpose[change | edit source]

Specific machines, special-purpose[change | edit source]