Design Principle of Memory Architecture of Parallel Processing Computer

Abstract

In modern time the focus of parallel computer development is being shifted from the processor perspective to the memory system perspective. If we will go to the Moore's law we will find the reason, which says that the performance processor in general doubled in each 18 months and in recent years the performance of memory is increasing by 8% per year which causes an increasing discrepancy between processor & memory. So this will cause that a processor must wait more & more for data from memory. So it becomes necessary to design a principle for memory architecture of parallel computer. Shared Memory, Distributed Memory, Distributed Shared Memory etc. are some memory architecture for parallel computer. But fast development of parallel computer gives lots of memory architecture. In this paper I give an extensive view of different memory architecture used in parallel computer, I try to give what are the basic requirements that the parallel processor needs from memory, according to this requirement that should be the architecture of memory. So this paper mainly covers what is the main principle on which of memory of parallel computer is designed.

Keywords: Shared: Shared Memory, Distributed Memory, Distributed Shared Memory