Accelerated Code Generator for Processing Ocean Color Remote Sensing Data on GPU
DC Field | Value | Language |
---|---|---|
dc.contributor.author | 허재무 | - |
dc.contributor.author | 조강원 | - |
dc.contributor.author | 한희정 | - |
dc.contributor.author | 양현 | - |
dc.date.accessioned | 2020-07-15T11:33:28Z | - |
dc.date.available | 2020-07-15T11:33:28Z | - |
dc.date.created | 2020-02-11 | - |
dc.date.issued | 2018-07-27 | - |
dc.identifier.uri | https://sciwatch.kiost.ac.kr/handle/2020.kiost/23179 | - |
dc.description.abstract | As the satellite data gradually increases, the processing time of the algorithm for remote sensing also increases. It is now essential to make efforts to improve the processing performance in various accelerators such as Graphics Processing Unit (GPU). However, it is very difficult to use programming models that make programs to run on accelerators. Especially for scientists, easy programming and performance are often more important than providing many optimization functionsand directives. To meet these requirements, we developed the accelerated code generator based on Open Computing Language (OpenCL) for processing ocean color remote sensing data. We easily applied our generator to atmospheric correction program for GOCI data as a case study and the program’s performance achieved about 29.2x that is as good as the hand-written OpenCL program. We look forward to many scientists that want to find a tool mentioned above taking advantage of our generator. | - |
dc.description.uri | 1 | - |
dc.language | English | - |
dc.publisher | GRSS/IEEE | - |
dc.relation.isPartOf | 2018 IGARSS | - |
dc.title | Accelerated Code Generator for Processing Ocean Color Remote Sensing Data on GPU | - |
dc.type | Conference | - |
dc.citation.endPage | 1 | - |
dc.citation.startPage | 1 | - |
dc.citation.title | 2018 IGARSS | - |
dc.contributor.alternativeName | 허재무 | - |
dc.contributor.alternativeName | 한희정 | - |
dc.contributor.alternativeName | 양현 | - |
dc.identifier.bibliographicCitation | 2018 IGARSS, pp.1 | - |
dc.description.journalClass | 1 | - |