Skip to main navigation Skip to search Skip to main content

Investigate the Computational Performance of Many-Core Graphics Processing Unit and Coprocessor for Fluid-Film Lubrication Analysis (Ii)

  • Wang, Nen-Zi (PI)

Project: National Science and Technology Council › National Science and Technology Council Academic Grants

Project Details

Abstract

This proposal is intended as an extended study of the current MOST project (MOST 104-2221-E-182-065, 2015/08-2016/07, which was proposed as a two-year project, but passed as a one-year project). Taking the notes from the reviewers’ comments (last year proposal) and the progress and results obtained in the current project, this project is to accomplish the enhanced objectives set forth by the original two-year proposal. The final results of this study will facilitate the publication of the research results due to a complete study of the subject. An efficient numerical computation for dealing with fluid-film lubrication analysis is an important research tool. In many cases numerical computations are required to obtain the solutions of various Reynolds equations and related equations, such as the equations considering the compressibility of the lubricant, the variation of viscosity, and the deformation of the bearing surfaces. As a result, the governing equations, such as in thermohydrodynamic lubrication, may include generalized Reynolds equation, energy equation, and elastic deformation equation. In recent years, the high performance computing is developed in a fast pace in many areas. These include multi-core CPUs/many-integrated-core (MIC) Coprocessor and graphics processing units (GPUs), as well as the open architecture of parallel programming paradigms, mainly OpenMP and OpenACC. The OpenACC standard is designed for programming in using GPUs, while the OpenMP is the de facto standard for processors with multiple processor cores, including multicore CPUs and MIC Coprocessors. In this project the computing capability of the multicore CPUs and the newly developed coprocessors from Intel is to be investigated using the OpenMP computing. The efficiency of the approach is to be assessed by using the numerical model for the steady-state and transient thermohydrodynamic lubrication analyses. And the effectiveness of the approach is to be compared with the current study of the OpenACC computing of the same lubrication model. The main objectives of the proposed project are: (1) Apply the OpenMP programming (for multicore CPUs and MIC Coprocessors) to solve the lubrication models to investigate the performance of the multithreaded computing as well as to develop the methods to increase the parallel efficiency; (2) Combine the computing power of the GPUs and coprocessors to further improve the performance of the numerical computation for fluid-film lubrication analysis.

Project IDs

Project ID:PB10507-2963
External Project ID:MOST105-2221-E182-019
StatusFinished
Effective start/end date01/08/16 → 31/07/17

Keywords

  • Thermohydrodynamic Lubrication
  • Generalized Reynolds Equation
  • Parallel Computing
  • Graphics

Fingerprint

Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.