Towards a clustered system for supercomputing

dc.contributor.author Lukyamuzi, Andrew
dc.date.accessioned 2012-09-27T11:20:23Z
dc.date.available 2012-09-27T11:20:23Z
dc.date.issued 2007-02
dc.description A Project Report submitted to School of Graduate Studies in partial fulfillment of the requirements for the award of Master of Science in Computer Science Degree of Makerere University. en_US
dc.description.abstract High Performance Clusters (HPCs) are increasingly becoming popular in the field of supercomputing. Improved performance/price ratio is the major driving force. As a result, HPCs are taking over the role of traditional super computers. This study investigates HPCs in relation to providing a cost effective solution to the problem of low processing power. It examines the available methodologies of implementing High Performance Computing using open source software. An evaluation of these methodologies is critically analyzed. The building blocks and challenges of HPCs are also given attention. Consequently, an openMOSIX cluster is proposed for adaptation in the Mathematics Department, Makerere University to enhance its computational power. The study also describes the structure and the design of the would be system. Finally, details of a mini cluster built using the proposed solution are given. en_US
dc.identifier.uri http://hdl.handle.net/10570/718
dc.language.iso en en_US
dc.publisher Makerere University en_US
dc.subject Cluster types en_US
dc.subject Supercomputing en_US
dc.subject Beowulf cluster en_US
dc.subject MOSIX cluster en_US
dc.title Towards a clustered system for supercomputing en_US
dc.type Thesis, masters en_US
Files
Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
lukyamuzi-andrew-cit-masters-report.pdf
Size:
1.13 MB
Format:
Adobe Portable Document Format
Description:
Report
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
1.71 KB
Format:
Item-specific license agreed upon to submission
Description: