Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Implementing Cauchy Reed-Solomon Utilizing OpenCL
Blekinge Institute of Technology, School of Computing.
2013 (English)Student thesisAlternative title
Cauchy Reed-Solomon implementerat med OpenCL (Swedish)
Abstract [en]

In this paper the performance of executing Cauchy Reed-Solomon (CRS) coding on the GPU is evaluated and compared with execution on the CPU. Three different prototypes are developed. One is sequential and developed in C++, the other two are developed utilizing C++ and OpenCL. The measurements are done by comparing the execution time of different data block sizes ranging from 16KB up to 256MB with two different encoding ratios, 9/12 and 10/16. The measurements are done on an Intel CPU with 4 cores with an integrated graphics card and an AMD graphics card. The OpenCL prototypes are executed with three different targets, the CPU, the integrated graphics card and the AMD graphics card. The sequential prototype is executed on the same CPU, but on a single core. The results from the measurements show that the GPU is faster than the CPU on larger data sizes. The highest measured throughput is achieved with the multithreaded CPU prototypes (OpenCL executed on the CPU) for sizes around 1MB.

Abstract [sv]

I den här rapporten genomförs prestandatester för exekvering av Cauchy Reed-Solomon (CRS) coding på grafikkortet och jämförs med exekvering på en CPU. Tre olika prototyper har utvecklats. En är sekventiell och utvecklad i C++, de två andra är utvecklade i OpenCL och C++. Testerna genomförs genom tidtagning på olika stora datablock, från 16KB upp till 256MB med olika enkodnings ratios, 9/12 och 10/16. CPU:n som används i testerna är en Intel CPU med 4 kärnor, och grafikkorten som används är det integrerade grafikkortet på CPU:n samt ett grafikkort från AMD. OpenCL prototyperna exekveras med tre olika inställningar, CPU för multitrådat, det integrerade Intel grafikkortet och det dedikerade AMD grafikkortet. Den sekventiella prototypen exekveras på samma CPU, men med en kärna. Resultaten från experimenten visar att grafikkorten är snabbare än CPU:n för större datablock. Den prototyp som fick högst genomströmning av data var den multitrådade CPU prototypen för datablock i storleksordningen 1MB.

Place, publisher, year, edition, pages
2013. , p. 40
Keywords [en]
OpenCL, Cauchy Reed-Solomon, GPGPU, Graphicscard
National Category
Mathematics Computer Sciences
Identifiers
URN: urn:nbn:se:bth-4823Local ID: oai:bth.se:arkivexBCB2FD16E55A96C2C1257C5E00666323OAI: oai:DiVA.org:bth-4823DiVA, id: diva2:832171
Educational program
PAACI Master of Science in Game and Software Engineering
Uppsok
Physics, Chemistry, Mathematics
Supervisors
Available from: 2015-04-22 Created: 2014-01-12 Last updated: 2018-01-11Bibliographically approved

Open Access in DiVA

fulltext(666 kB)193 downloads
File information
File name FULLTEXT01.pdfFile size 666 kBChecksum SHA-512
801b0d3677c47259e64484dcb3488e50dc5ea58a562d5489f0fdb4a886880180841130871f2a9da9a0bfa5b3fd5e5c3264186849aeeab9089c4e9d41dd078773
Type fulltextMimetype application/pdf

By organisation
School of Computing
MathematicsComputer Sciences

Search outside of DiVA

GoogleGoogle Scholar
Total: 193 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 821 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf