cuIBM -- A GPU-accelerated Immersed Boundary Method

2011/09/16 by Simon K. Layton, Simon K Layton, Anush Krishnan +4
Computer Science · Engineering · #Computational Engineering #FOS: Computer and information sciences #Finance #Fluid Dynamics Simulations and Interactions #Fluid Dynamics and Heat Transfer #Lattice Boltzmann Simulation Studies #and Science (cs.CE) #cs.CE

paper · pdf · doi:10.48550/arxiv.1109.3524

Extended paper post-conference, presented at the 23rd International Conference on Parallel Computational Fluid Dynamics (http://www.parcfd.org), ParCFD 2011, Barcelona (unpublished)

openalex publication_date 2011/09/16 · arxiv created 2016/04/09 · arxiv updated 2016/04/12 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

A projection-based immersed boundary method is dominated by sparse linear algebra routines. Using the open-source Cusp library, we observe a speedup (with respect to a single CPU core) which reflects the constraints of a bandwidth-dominated problem on the GPU. Nevertheless, GPUs offer the capacity to solve large problems on commodity hardware. This work includes validation and a convergence study of the GPU-accelerated IBM, and various optimizations.

Citations

• Numerical solution of the Navier-Stokes equations

Related

• Simple and efficient GPU parallelization of existing H-Matrix accelerated BEM code
• Efficient LBM on GPUs for dense moving objects using immersed boundary condition
• Explicit caching HYB: a new high-performance SpMV framework on GPGPU
• Condensed interior-point methods: porting reduced-space approaches on GPU hardware
• Condensed Interior-Point Methods for Scalable Nonlinear Programming on GPUs