GridComputing

Last edit August 14, 2008

A computing model where computational resources (processing and/or storage) are coupled into a parallel and distributed system that spans multiple geographic and administrative domains.

Grid computing is related to PeerToPeer systems and clustering solutions. It is different from PeerToPeer in that it includes hardware sharing. It is different from clustering, since clustering mostly works on nodes of identical hardware and operating systems, in a centralized manner.

The GlobalGridForum is publishing documents in a W3C manner. See http://www.ggf.org

Globus has developed a set of standards and some tools or API's. (http://www.globus.org). The Globus stuff is used by IBM as the foundation for their GridComputing initiatives.

IBM has a so-called redbook on GridComputing, an introduction to the concepts of GridComputing here: http://publib-b.boulder.ibm.com/Redbooks.nsf/RedbookAbstracts/redp3613.html?Open

Oracle has been working on Grid computing as well. They point out that using cheap computers is a viable solution as opposed to IBM who does not do so. http://otn.oracle.com/oramag/webcolumns/2003/opinion/souder_grid.html

Contibutors: PieterVerbaarschott, MichaelFinney. (add your name here to avoid ThreadMode)