Computing of discretized reynolds equations using graphics processing unit

High-performance graphics processing units (GPU) computing is an emerging powerful tool for engineering and scientific modeling. The main advantage of using GPU computing is its ability to handle a large number of concurrent threads. The solution methods for the discretized incompressible- and compressible-fluid Reynolds equations suitable for use in a workstation with a GPU were studied. The lubrication models used were an inclined-surface slider with or without a central recess and an air journal bearing. The performance of the GPU computing was close to the multithreaded computing in which the grid sizes are large. The performance of the Accelerator Programming Model of Portland Group programming was compatible with Compute Unified Device Architecture computing in the slider cases, but was significantly worse for air bearings. The approach should be useful for lubrication engineers using workstations equipped with multi-core CPU and many-core GPU. This is an abstract of a paper presented at the Society of Tribologists & Lubrication Engineers Annual Meeting and Exhibition 2012 (St. Louis, MO, 5/6-10/2012).