Research Symposium >
SLIIT Student Research -2017 >

Please use this identifier to cite or link to this item:

Title: P2P GPGPU Distributed Computing Framework
Authors: Wijerathna, I.G.A.S.
Hariprasanth, S.
Hewage, M.H.S.M.
Ranathunge, T.R.
Kodagoda, Nuwan
Keywords: GPGPU computing
Distributed computing parallel algorithms
peer-to-peer network
Fault toleranc
Machine learning
Issue Date: 2017
Publisher: SLIIT
Series/Report no.: ;17-068
Abstract: Abstract—In this present technological world, we can see a rapid revolutionary development in the field of IT. As the developments imply as well as the complexity of the problems that the computers can also solve gradually increases. Hence, Advancements of technologies and more computational power is essential to complete complex algorithms. Therefore, such kinds of algorithms need large servers. But we can’t anticipate such powerful resources are available to everyone. If there is a possibility to share its resources with another computer, it makes an immense revolution. In this paper, we proposed a peer-to-peer network that allows sharing idle processing power among other peers who are connected to the existing network. Hence, the shared processing power can be used to execute a complex problem by distributing it. The intention of this process to speed up the ordinary problem-solving procedure by parallel algorithms and parallelized data. In such scenario, proper scheduling and data splitting mechanism can improve the performance, cost of computation, and increase the reliability of the system. The application is based on a peer-to-peer network since there is a high probability of occurring faults. To reduce these errors, the system is implemented with fault tolerance mechanisms.
Appears in Collections:SLIIT Student Research -2017

Files in This Item:

File Description SizeFormat
17-068.pdf850.36 kBAdobe PDFView/Open
View Statistics

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.


Valid XHTML 1.0! DSpace Software Copyright © 2002-2010  Duraspace - Feedback