Show simple item record

dc.contributor.authorBurge, Legand L. III
dc.date.accessioned2014-10-01T13:33:23Z
dc.date.available2014-10-01T13:33:23Z
dc.date.issued1995-07-01
dc.identifier.urihttps://hdl.handle.net/11244/12698
dc.description.abstractA sequential computer executes one CPU instruction at a time. Over the years sequential computers have increased steadily in performance primarily as a result of improvements in digital hardware technology. One major concern of computer designers is that logic and memory devices are approaching ultimate physical limits on their size and speed. While size reductions and speed increases of a few orders of magnitude beyond present levels seem feasible, further improvements in the performance of sequential computers may not be achievable at acceptable cost. A more economic solution is to design systems that can process more than one CPU instruction at a time. This is known as parallel processing. Parallel processors are also referred to as distributed systems. These systems consists of an interconnected collection of autonomous computers [Sta84]. There are many ways of classifying distributed systems based on their structure or behavior. Based on Flynn's taxonomy of computer architectures, distributed systems belong to the MIMD (multiple instruction multiple data) class of computer architectures [Tan92]. The MIMD class consist of two categories: those that have shared memory (tightly coupled), and those that do not (loosely coupled). As shown in Figure 1.1, each category can be furthered divided based on the architecture of the interconnection network. In bus-based systems, there is a single network, backplane, bus, cable, or other medium that connects all machines. Switched systems connect machines by individual wires.
dc.formatapplication/pdf
dc.languageen_US
dc.publisherOklahoma State University
dc.rightsCopyright is held by the author who has granted the Oklahoma State University Library the non-exclusive right to share this material in its institutional repository. Contact Digital Library Services at lib-dls@okstate.edu or 405-744-9161 for the permission policy on the use, reproduction or distribution of this material.
dc.titleDecentralized Algorithm for Communication Efficient Distributed Shared Memory
dc.typetext
osu.filenameThesis-1995-B954d.pdf
osu.accesstypeOpen Access
dc.type.genreThesis


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record