الفهرس Only 14 pages are availabe for public view
 |  | from | 207 |  |  |
| | | from | 207 | | |
Abstract
IN 1980S, IT WAS BELIVED THAT COMPUTER PERFORMANCE COULD BE BEST IMPROVED BY CREATING FASTER AND MORE EFFICIENT PROCESSORS. THIS IDEA WAS CHALLENGED BY PARALLEL PROCESSING, WHICH MEANS LINKING A GROUP OF COMPUTERS TO SOLVE A COMPUTATIONAL PROBLEM. SINCE THE EEATLY 1990S THERE HAS BEEN AN INCREASING TREND TREND FOR MOVING from EXPENSIVE AND SPECIAL PARALLEL SUPERCOMPUTERS TOWARDS NETWORKS OF PCS. THIS IS CALLED A CLUSTER OF COMPUTERS. THESE SYSTEMS ARE SCALABLE, I.E, THEY CAN BE TUNED TO AVAILABLE BUDGET AND COMPUTATIONAL NEEDS. ONE OF THE FACTORS THAT HAVE ENABLED THIS TRANSITION HAS BEEN THE RAPID IMPROVEMENT IN THE AVAILABILITY OF COMMODITY HIGH PERFORMANCE COMPOENTS FOR PCS AND NETWORKS. APPLICATIONS THAT CAN BE PARALLELIZED BY DISTRIBUTING ACROSS THE PROCESSORS USE EITHER MESSAGE PASSING OR SHARED MEMORY FOR COMMUNICATION.