This paper presents a compression scheme for large point scans including per-point normals. For the encoding of such scans, the paper introduces a type of closest sphere packing grids, the hexagonal close packing (HCP). To compress the data, linear sequences of filled cells in HCP grids are extracted. Point positions and normals in these runs are incrementally encoded. At a grid spacing close to the point sampling distance, the compression scheme only requires slightly more than 3 bits per point position. Incrementally encoded per-point normals are quantized at high fidelity using only 5 bits per normal. The compressed data stream can be decoded in the graphics processing unit (GPU). Decoded point positions are saved in graphics memory, and they are then used on the GPU again to render point primitives. In this way gigantic point scans are rendered from their compressed representation in local GPU memory at interactive frame rates. (http://wwwcg.in.tum.de/Research/data/Publications/pbg05.pdf)

As simulation and rendering capabilities continue to increase, volumetric effects like smoke, fire or explosions will be frequently encountered in computer games and virtual environments. This paper presents techniques for the visual simulation and rendering of such effects that keep up with the demands for frame rates imposed by such environments. This is achieved by leveraging functionality on recent graphics programming units (GPUs) in combination with a novel approach to model non physics-based, yet realistic variations in flow fields. The paper shows how to use this mechanism for simulating effects. Physics-based simulation is performed on 2D proxy geometries, and simulation results are extruded to 3D using particle or texture based approaches. (http://wwwcg.in.tum.de/Research/data/Publications/eg05.pdf)

This paper presents a particle system for interactive visualization of steady 3D flow fields on uniform grids. For large particle systems, particle integration needs to be accelerated and the transfer of particle data to the GPU must be avoided. To fulfill these requirements, this paper exploits features of recent graphics accelerators to advect particles in the graphics processing unit (GPU), saving particle positions in graphics memory, and then sending these positions through the GPU again to obtain images in the frame buffer. (http://wwwcg.in.tum.de/Research/data/Publications/tvcg05.pdf)

Parallel genetic algorithms are usually implemented on parallel machines or distributed systems. This paper describes how fine-grained parallel genetic algorithms can be mapped to programmable graphics hardware found in commodity PCs. The approach stores chromosomes and their fitness values in texture memory on the graphics card. Both fitness evaluation and genetic operations are implemented entirely with fragment programs executed on the GPU in parallel. The paper demonstrate the effectiveness of this approach by comparing it with a compatible software implementation. The presented approach benefits from the advantages of parallel genetic algorithms on a low-cost platform. (http://www.cad.zju.edu.cn/home/yqz/)

This VMV 2004 paper by De Chiara et al. presents a massive simulation of a behavioral model using graphics hardware. A well-known flocking model is implemented on the GPU. The model is capable of managing large aggregate motion of birds in a virtual environment including avoidance of both static and dynamic obstacles. The effectiveness of the GPU implementation is demonstrated with a comparison to a CPU implementation. (Massive Simulation using GPU of a distributed behavioral model of a flock with obstacle avoidance. Rosario De Chiara, Ugo Erra, Vittorio Scarano, Maurizio Tatafiore. In Proceedings of 9th Internation Fall Workshop VISION, MODELLING, AND VISUALIZATION 2004.)

Graphics hardware’s rapid evolving pace has made self-adaptable software very desirable. Changhao Jiang and Marc Snir at University of Illinois Urbana Champaign have developed a library generator for graphics hardware, that can automatically generate high performance matrix multiplication with comparable performance to expert manually tuned version on various graphics hardware platforms. The paper will be published at the Fourteenth International Conference on Parallel Architecture and Compilation Techniques (PACT) 2005. (Automatic Tuning Matrix Multiplication on Graphics Hardware)

From the abstract: In recent years, the development of programmable graphics pipelines has placed the power of parallel computation in the hands of consumers. Systems developers are now paying attention to the general purpose computational ability of these graphics processor units, or GPUs, and are using them in novel ways. This paper examines using pixel shaders for executing audio algorithms. We compare GPU performance to CPU performance, discuss problems encountered, and suggest new directions for supporting the needs of the audio community. Source code is also available. (Audio and the Graphics Processing Unit”, by Sean Whalen)

This paper by Chu and Tai at HKUST presents a physically-based method for simulating ink dispersion in absorbent paper for art creation purposes. The ink flow model is based on the lattice Boltzmann equation and is designed to work on the GPU efficiently. (MoXi: Real-Time Ink Dispersion in Absorbent Paper. Nelson S.-H. Chu and Chiew-Lan Tai. To appear in ACM Transactions on Graphics (SIGGRAPH 2005 issue), August 2005)

The task of reconstructing an object from its projections via tomographic methods is a time-consuming process due to the vast complexity of the data. GPUs offer an affordable alternative to proprietary ASICs and FPGAs. Fang Xu and Klaus Mueller at Stony Brook University have shown that the latest generation of GPUs can be exploited to perform both analytical and iterative reconstruction from X-ray and functional imaging data at clinical rates and high quality. Visualization of the reconstructed object is easily achieved since the object already resides in the graphics hardware, allowing one to run a visualization module at any time to view the reconstruction results. Their implementation allows speedups of 1-2 orders of magnitude over software implementations, at comparable image quality. (Link to the project page)

This work is part of a Masters thesis by Konrad Pietras of Technical University of Lodz, Poland. The method uses a neural network, implemented in a vertex program, for approximating the light model described in “Display of The Earth Taking into Account Atmospheric Scattering” by Nishita et al., SIGGRAPH 1996. (GPU-based perceptron used for 4-dimensional texture lookup)