Page Banner

United States Department of Agriculture

Agricultural Research Service

Research Project: IMPROVING COMPUTATIONAL MODELING IN SUPPORT OF BETTER EROSION AND SEDIMENT MOVEMENT CONTROL IN AGRICULTURAL WATERSHEDS

Location: Watershed Physical Processes Research Unit

Title: Parallelized CCHE2D flow model with CUDA Fortran on Graphics Process Units

Authors
item Zhang, Y -
item Jia, Y -

Submitted to: Computers & Fluids
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: June 1, 2013
Publication Date: July 1, 2013
Citation: Zhang, Y., Jia, Y. 2013. Parallelized CCHE2D flow model with CUDA Fortran on Graphics Process Units. Computers & Fluids. 84:359-368. Available: www.journals.elsevier.com/computers-and-fluids/

Interpretive Summary: This paper presents the CCHE2D implicit flow model parallelized using CUDA Fortran programming technique on Graphics Processing Units (GPUs). A parallelized implicit Alternating Direction Implicit (ADI) solver using Parallel Cyclic Reduction (PCR) algorithm on GPU is developed and tested. This solver outperforms the Strong Implicit Procedure (SIP) solver and its parallel alternatives. Computing accuracy and efficiency of both CPU and GPU versions of models were compared with one experimental case and one field case. It has been demonstrated that the parallelized CCHE2D flow model with CUDA Fortran is capable of accurately predicting steady flow or unsteady flow with a much higher computing efficiency on the GPU. The parallelized CCHE2D model was validated using one experimental case, which has proved its consistency and higher efficiency compared to the serial version. The parallelized CCHE2D model has also been successfully applied to a long term, large scale unsteady flow simulation. Satisfactory results on a series of refined meshes were obtained in significantly shorter time.

Technical Abstract: This paper presents the CCHE2D implicit flow model parallelized using CUDA Fortran programming technique on Graphics Processing Units (GPUs). A parallelized implicit Alternating Direction Implicit (ADI) solver using Parallel Cyclic Reduction (PCR) algorithm on GPU is developed and tested. This solver outperforms the Strong Implicit Procedure (SIP) solver and its parallel alternatives. Computing accuracy and efficiency of both CPU and GPU versions of models were compared with one experimental case and one field case. It has been demonstrated that the parallelized CCHE2D flow model with CUDA Fortran is capable of accurately predicting steady flow or unsteady flow with a much higher computing efficiency on the GPU.

Last Modified: 10/22/2014
Footer Content Back to Top of Page