11876

An Approach to Efficient FEM Simulations on Graphics Processing Units Using CUDA

Bjorn Nutti, Dragan Marinkovic
AlgoritmFabriken AB, Stockholm, Sweden
Facta Universitatis, Series: Mechanical Engineering Vol. 12, No 1, pp. 15 – 25, 2014
@article{nutti2014approach,

   title={AN APPROACH TO EFFICIENT FEM SIMULATIONS ON GRAPHICS PROCESSING UNITS USING CUDA},

   author={Nutti, Bj{"o}rn and Marinkovi{‘c}, Dragan},

   journal={Facta Universitatis, Series: Mechanical Engineering},

   volume={12},

   number={1},

   pages={15–25},

   year={2014}

}

Download Download (PDF)   View View   Source Source   

1426

views

The paper presents a highly efficient way of simulating the dynamic behavior of deformable objects by means of the finite element method (FEM) with computations performed on Graphics Processing Units (GPU). The presented implementation reduces bottlenecks related to memory accesses by grouping the necessary data per node pairs, in contrast to the classical way done per element. This strategy reduces the memory access patterns that are not suitable for the GPU memory architecture. Furthermore, the presented implementation takes advantage of the underlying sparse-block-matrix structure, and it has been demonstrated how to avoid potential bottlenecks in the algorithm. To achieve plausible deformational behavior for large local rotations, the objects are modeled by means of a simplified co-rotational FEM formulation.
VN:F [1.9.22_1171]
Rating: 5.0/5 (3 votes cast)
An Approach to Efficient FEM Simulations on Graphics Processing Units Using CUDA, 5.0 out of 5 based on 3 ratings

* * *

* * *

Follow us on Twitter

HGPU group

1943 peoples are following HGPU @twitter

Like us on Facebook

HGPU group

442 people like HGPU on Facebook

HGPU group © 2010-2016 hgpu.org

All rights belong to the respective authors

Contact us: