Dipl.-Inf. Marco Ament

VISUS - Visualization Research Center
Universität Stuttgart
Nobelstraße 15
D-70569 Stuttgart, Germany

Room: TTI, 0.66
Email: marco.ament [at] vis.uni-stuttgart.de
Phone: +49 7117816185
www: http://www.vis.uni-stuttgart.de/~amentmo/

Research:

A Parallel Preconditioned Conjugate Gradient Solver for the Poisson Problem on a Multi-GPU platform M. Ament, G. Knittel, D. Weiskopf, W. Straßer To appear at the 18th Euromicro International Conference on Parallel, Distributed and Network-Based Computing (PDP2010), 2010 Special Session: Parallel Algorithms and Software for Sparse Linear Algebra Computations Download PDF

Abstract: We present a parallel conjugate gradient solver for the Poisson problem optimized for multi-GPU platforms. Our approach includes a novel heuristic Poisson preconditioner which is well-suited for massively-parallel SIMD processing. Furthermore, we address the problem of limited transfer rates over typical data channels such as the PCI-express bus relative to the bandwidth requirements of powerful GPUs. Specifically, naïve communication schemes can severely reduce the achievable speedup in such communication-intense algorithms. For this reason, we employ overlapping memory transfers to establish a high level of concurrency and to improve scalability. We have implemented our model on a high-performance workstation with multiple hardware accelerators. We will discuss the mathematical principles, give implementation details, and present the performance and the scalability of the system.

Dynamic Grid Refinement for Fluid Simulations on Parallel Graphics Architectures M. Ament, W. Straßer EUROGRAPHICS Symposium on Parallel Graphics and Visualization (EGPGV09), 2009 Download PDF Project page (images + movies)

Abstract: We present a physically-based fluid simulation with dynamic grid refinement on parallel SIMD graphics hardware. The irregular and dynamic structure of an adaptive grid requires sophisticated memory access patterns as well as a decomposition of the problem for parallel processing and the distribution of tasks to multiple threads. In this paper, we focus on the representation and management of the dynamic grid on the graphics device for an efficient parallelization of the advection step and the iterative solving of the Poisson equation. In order to achieve high performance, we utilize the hardware’s capabilities like fast cache access and trilinear filtering. Furthermore, expensive data transfer between host and device is minimized to avoid a major bottleneck. We report results on the inherent overhead of the dynamic grid compared to an equivalent Cartesian grid. In addition, a visual simulation of smoke is presented with radiosity-based illumination and volume ray casting at interactive frame rates.

Hardware Accelerated Fluid Dynamics with Adaptive Grid Refinement M. Ament WSI/GRIS, University of Tübingen, Diploma Thesis, 2008 Download PDF

Abstract: In this thesis, a physically-based fluid simulation with dynamic grid refinement parallel SIMD graphics hardware is presented. The irregular and dynamic structure of an adaptive grid requires sophisticated memory access patterns as well as a decomposition of the problem for parallel processing and the distribution of tasks to multiple threads. The focus of this thesis lies on the representation and management of the dynamic grid on the graphics device for an efficient parallelization of the advection step and the iterative solving of the Poisson equation. In order to achieve high performance, the hardware's capabilities like fast cache access and trilinear filtering are utilized. Furthermore, expensive data transfer between host and device is minimized to avoid a major bottleneck. Results on the inherent overhead of the dynamic grid compared to an equivalent Cartesian grid are reported. In addition, a visual simulation of smoke is presented with radiosity-based illumination and volume ray casting at interactive frame rates.