Particle Image Velocimetry Data Processing On A Gpu Cluster
| dc.contributor.author | Dallas, Cameron | |
| dc.contributor.author | Sullivan, Pierre | |
| dc.date.accessioned | 2018-11-07T00:45:09Z | |
| dc.date.available | 2018-11-07T00:45:09Z | |
| dc.date.issued | May-18 | |
| dc.description | Paper presented at 2018 Canadian Society of Mechanical Engineers International Congress, 27-30 May 2018. | en_US |
| dc.description.abstract | Particle image velocimetry (PIV) data processing is a computationally expensive process. The immense time taken to analyze data can limit the maximum dataset size. Using graphics processing units (GPUs) has been shown to drastically decrease the processing time for PIV image pairs. The open-source PIV data processing software OpenPIV has been ported to run on a GPU to boost speed and efficiency and has outperformed the CPU version of the software. A multipass method is being implemented in OpenPIV to improve both speed and accuracy. The completed algorithm will be tested on an embedder CPU-GPU device, a desktop computer, and the SOSCIP GPU-accelerated supercomputing cluster. Ultimately, OpenPIV will run on a wide variety of computer platforms an enable larger datasets to be collects, leading to better statistics on the resulting velocity fields. | en_US |
| dc.identifier | CSME119 | |
| dc.identifier.issn | 978-1-77355-023-7 | |
| dc.identifier.uri | http://hdl.handle.net/10315/35287 | |
| dc.identifier.uri | http://dx.doi.org/10.25071/10315/35287 | |
| dc.language.iso | en | en_US |
| dc.publisher | CSME-SCGM | en_US |
| dc.rights | The copyright for the paper content remains with the author. | |
| dc.subject | PIV | |
| dc.subject | GPU | |
| dc.subject | Fluid Mechanics | en_US |
| dc.title | Particle Image Velocimetry Data Processing On A Gpu Cluster | en_US |
| dc.type | Article | en_US |