Current Graphics Processing Units Research Articles

Optimizing H.264/AVC interprediction on a GPU‐based framework

SUMMARYH.264/MPEG‐4 part 10 is the latest standard for video compression and promises a significant advance in terms of quality and distortion compared with the commercial standards currently most in use such as MPEG‐2 or MPEG‐4. To achieve this better performance, H.264 adopts a large number of new/improved compression techniques compared with previous standards, albeit at the expense of higher computational complexity. In addition, in recent years new hardware accelerators have emerged, such as graphics processing units (GPUs), which provide a new opportunity to reduce complexity for a large variety of algorithms. However, current GPUs suffer from higher power consumption requirements because of its design. Up to now, GPU‐based software developers have not taken this into account. In this paper, we present a detailed procedure to implement the H.264 motion estimation for a GPU, with the aim of reducing time and, as a consequence, the energy consumption. The results show a negligible drop in rate distortion with a time reduction of over 91.5% on average and it reduces the energy consumption by a factor of 11.78 compared with the reference implementation. Copyright © 2011 John Wiley & Sons, Ltd.

EIT Forward Problem Parallel Simulation Environment with Anisotropic Tissue and Realistic Electrode Models

Electrical impedance tomography (EIT) is an imaging technology based on impedance measurements. To retrieve meaningful insights from these measurements, EIT relies on detailed knowledge of the underlying electrical properties of the body. This is obtained from numerical models of current flows therein. The nonhomogeneous and anisotropic electric properties of human tissues make accurate modeling and simulation very challenging, leading to a tradeoff between physical accuracy and technical feasibility, which at present severely limits the capabilities of EIT. This work presents a complete algorithmic flow for an accurate EIT modeling environment featuring high anatomical fidelity with a spatial resolution equal to that provided by an MRI and a novel realistic complete electrode model implementation. At the same time, we demonstrate that current graphics processing unit (GPU)-based platforms provide enough computational power that a domain discretized with five million voxels can be numerically modeled in about 30 s.

Fine-sorting one-dimensional particle-in-cell algorithm with Monte-Carlo collisions on a graphics processing unit

Particle-in-cell (PIC) simulations with Monte-Carlo collisions are used in plasma science to explore a variety of kinetic effects. One major problem is the long run-time of such simulations. Even on modern computer systems, PIC codes take a considerable amount of time for convergence. Most of the computations can be massively parallelized, since particles behave independently of each other within one time step. Current graphics processing units (GPUs) offer an attractive means for execution of the parallelized code. In this contribution we show a one-dimensional PIC code running on NVIDIA® GPUs using the CUDA™ environment. A distinctive feature of the code is that size of the cells that the code uses to sort the particles with respect to their coordinates is comparable to size of the grid cells used for discretization of the electric field. Hence, we call the corresponding algorithm “fine-sorting”. Implementation details and optimization of the code are discussed and the speed-up compared to classical CPU approaches is computed.

Sponge

Graphics processing units (GPUs) provide a low cost platform for accelerating high performance computations. The introduction of new programming languages, such as CUDA and OpenCL, makes GPU programming attractive to a wide variety of programmers. However, programming GPUs is still a cumbersome task for two primary reasons: tedious performance optimizations and lack of portability. First, optimizing an algorithm for a specific GPU is a time-consuming task that requires a thorough understanding of both the algorithm and the underlying hardware. Unoptimized CUDA programs typically only achieve a small fraction of the peak GPU performance. Second, GPU code lacks efficient portability as code written for one GPU can be inefficient when executed on another. Moving code from one GPU to another while maintaining the desired performance is a non-trivial task often requiring significant modifications to account for the hardware differences. In this work, we propose Sponge, a compilation framework for GPUs using synchronous data flow streaming languages. Sponge is capable of performing a wide variety of optimizations to generate efficient code for graphics engines. Sponge alleviates the problems associated with current GPU programming methods by providing portability across different generations of GPUs and CPUs, and a better abstraction of the hardware details, such as the memory hierarchy and threading model. Using streaming, we provide a write-once software paradigm and rely on the compiler to automatically create optimized CUDA code for a wide variety of GPU targets. Sponge's compiler optimizations improve the performance of the baseline CUDA implementations by an average of 3.2x.

Sponge

Graphics processing units (GPUs) provide a low cost platform for accelerating high performance computations. The introduction of new programming languages, such as CUDA and OpenCL, makes GPU programming attractive to a wide variety of programmers. However, programming GPUs is still a cumbersome task for two primary reasons: tedious performance optimizations and lack of portability. First, optimizing an algorithm for a specific GPU is a time-consuming task that requires a thorough understanding of both the algorithm and the underlying hardware. Unoptimized CUDA programs typically only achieve a small fraction of the peak GPU performance. Second, GPU code lacks efficient portability as code written for one GPU can be inefficient when executed on another. Moving code from one GPU to another while maintaining the desired performance is a non-trivial task often requiring significant modifications to account for the hardware differences. In this work, we propose Sponge, a compilation framework for GPUs using synchronous data flow streaming languages. Sponge is capable of performing a wide variety of optimizations to generate efficient code for graphics engines. Sponge alleviates the problems associated with current GPU programming methods by providing portability across different generations of GPUs and CPUs, and a better abstraction of the hardware details, such as the memory hierarchy and threading model. Using streaming, we provide a write-once software paradigm and rely on the compiler to automatically create optimized CUDA code for a wide variety of GPU targets. Sponge's compiler optimizations improve the performance of the baseline CUDA implementations by an average of 3.2x.

Load Balancing versus Occupancy Maximization on Graphics Processing Units: The Generalized Hough Transform as a Case Study

Programs developed under the Compute Unified Device Architecture obtain the highest performance rate, when the exploitation of hardware resources on a Graphics Processing Unit (GPU) is maximized. In order to achieve this purpose, load balancing among threads and a high value of processor occupancy, i.e. the ratio of active threads, are indispensable. However, in certain applications, an optimally balanced implementation may limit the occupancy, due to a greater need for registers and shared memory. This is the case of the Fast Generalized Hough Transform (Fast GHT), an image-processing technique for localizing an object within an image. In this work, we present two parallelization alternatives for the Fast GHT, one that optimizes the load balancing and another that maximizes the occupancy. We have compared them using a large amount of real images to test their strong and weak points and we have drawn several conclusions about under which conditions it is better to use one or the other. We have also tackled several parallelization problems related to sparse data distribution, divergent execution paths, and irregular memory access patterns in updating operations by proposing a set of generic techniques, including compacting, sorting, and memory storage replication. Finally, we have compared our Fast GHT with the classic GHT, both on a current GPU, obtaining an important speed-up.

Simulation and visualization of the Saint-Venant system using GPUs

We consider three high-resolution schemes for computing shallow-water waves as described by the Saint-Venant system and discuss how to develop highly efficient implementations using graphical processing units (GPUs). The schemes are well-balanced for lake-at-rest problems, handle dry states, and support linear friction models. The first two schemes handle dry states by switching variables in the reconstruction step, so that bilinear reconstructions are computed using physical variables for small water depths and conserved variables elsewhere. In the third scheme, reconstructed slopes are modified in cells containing dry zones to ensure non-negative values at integration points. We discuss how single and double-precision arithmetics affect accuracy and efficiency, scalability and resource utilization for our implementations, and demonstrate that all three schemes map very well to current GPU hardware. We have also implemented direct and close-to-photo-realistic visualization of simulation results on the GPU, giving visual simulations with interactive speeds for reasonably-sized grids.

Molecular dynamics simulations of aqueous ions at the liquid–vapor interface accelerated using graphics processors

Molecular dynamics (MD) simulations are a vital tool in chemical research, as they are able to provide an atomistic view of chemical systems and processes that is not obtainable through experiment. However, large-scale MD simulations require access to multicore clusters or supercomputers that are not always available to all researchers. Recently, scientists have returned to exploring the power of graphics processing units (GPUs) for various applications, such as MD, enabled by the recent advances in hardware and integrated programming interfaces such as NVIDIA's CUDA platform. One area of particular interest within the context of chemical applications is that of aqueous interfaces, the salt solutions of which have found application as model systems for studying atmospheric process as well as physical behaviors such as the Hoffmeister effect. Here, we present results of GPU-accelerated simulations of the liquid-vapor interface of aqueous sodium iodide solutions. Analysis of various properties, such as density and surface tension, demonstrates that our model is consistent with previous studies of similar systems. In particular, we find that the current combination of water and ion force fields coupled with the ability to simulate surfaces of differing area enabled by GPU hardware is able to reproduce the experimental trend of increasing salt solution surface tension relative to pure water. In terms of performance, our GPU implementation performs equivalent to CHARMM running on 21 CPUs. Finally, we address possible issues with the accuracy of MD simulaions caused by nonstandard single-precision arithmetic implemented on current GPUs.

FAST RECONSTRUCTION METHOD BASED ON COMMON UNIFIED DEVICE ARCHITECTURE (CUDA) FOR MICRO-CT

Three-dimensional image reconstruction with Feldkamp, Davis, and Kress (FDK) algorithm is the most time consuming part in Micro-CT. The parallel algorithm based on the computer cluster is capable of accelerating image reconstruction speed; however, the hardware is very expensive. In this paper, using the most current graphics processing units (GPU), we present a method based on common unified device architecture (CUDA) for speeding up the Micro-CT image reconstruction process. The most time consuming filtering and back-projection parts of the FDK algorithm are parallelized for the CUDA architecture. The CUDA-based reconstruction speed and image qualities are compared with CPU results for the projecting data of the Micro-CT system. The results show that the 3D image reconstruction speed based on CUDA is ten times faster than the speed with CPU. In conclusion the FDK algorithm based on CUDA for Micro-CT can reconstruct the 3D image right after the end of data acquisition.

Semi‐Uniform Adaptive Patch Tessellation

Abstract We present an adaptive tessellation scheme for surfaces consisting of parametric patches. The resulting tessellations are topologically uniform, yet consistent and watertight across boundaries of patches with different tessellation levels. Our scheme is simple to implement, requires little memory and is well suited for instancing, a feature available on current Graphical Processing Units that allows a substantial performance increase. We describe how the scheme can be implemented efficiently and give performance benchmarks comparing it to some other approaches.

Diffusion tensor fiber tracking on graphics processing units

Diffusion tensor magnetic resonance imaging has been successfully applied to the process of fiber tracking, which determines the location of fiber bundles within the human brain. This process, however, can be quite lengthy when run on a regular workstation. We present a means of executing this process by making use of the graphics processing units of computers’ video cards, which provide a low-cost parallel execution environment that algorithms like fiber tracking can benefit from. With this method we have achieved performance gains varying from 14 to 40 times on common computers. Because of accuracy issues inherent to current graphics processing units, we define a variation index in order to assess how close the results obtained with our method are to those generated by programs running on the central processing units of computers. This index shows that results produced by our method are acceptable when compared to those of traditional programs.

Real-time visualization of large volume datasets on standard PC hardware

In medical area, interactive three-dimensional volume visualization of large volume datasets is a challenging task. One of the major challenges in graphics processing unit (GPU)-based volume rendering algorithms is the limited size of texture memory imposed by current GPU architecture. We attempt to overcome this limitation by rendering only visible parts of large CT datasets. In this paper, we present an efficient, high-quality volume rendering algorithm using GPUs for rendering large CT datasets at interactive frame rates on standard PC hardware. We subdivide the volume dataset into uniform sized blocks and take advantage of combinations of early ray termination, empty-space skipping and visibility culling to accelerate the whole rendering process and render visible parts of volume data. We have implemented our volume rendering algorithm for a large volume data of 512 × 304 × 1878 dimensions (visible female), and achieved real-time performance (i.e., 3–4 frames per second) on a Pentium 4 2.4 GHz PC equipped with NVIDIA Geforce 6600 graphics card (256 MB video memory). This method can be used as a 3D visualization tool of large CT datasets for doctors or radiologists.

Interactive Diffusion-based Smoothing and Segmentation of Volumetric Datasets on Graphics Hardware

Volume segmentation with concurrent visualization is becoming an increasingly important part of medical diagnostics. This is due to the fact that the immediate visual feedback speeds up evaluation of the segmentation process, hence enhances segmentation quality. Therefore, our aim was to develop a method for volume segmentation and smoothing which achieves interactive performance on standard PCs and is useful in clinical practice (i.e. fast and of high quality). Our application is based on seeded region growing and nonlinear isotropic as well as anisotropic diffusion. We use current GPUs (graphics processing units) to speed up the computation of the diffusion process and use hardware-accelerated interactive volume rendering. Using our approach the user can observe the diffusion process in real-time, change parameters interactively and view the result in a high-quality 3D direct volume rendering (DVR). The interactive nature of our algorithm and simultaneous visualization improved the usability of our segmentation and smoothing algorithm and proved useful in the clinical workflow. Using our application we were able to speed up the (an)isotropic diffusion process to achieve interactive performance.

Using graphics boards to compute holograms

Today's,consumer graphics boards incorporate highly integrated, parallel-working graphics processing units (GPUs) with transistor counts and performance that exceed those of CPUs. In addition to playing the latest 3D video game, you can use the graphics board's power to solve computational problems in science or engineering work. Current GPUs are programmable and flexible enough to transfer computational problems from the CPU to the GPU.Why shift computational work to the GPU? The short answer is processing power. As we demonstrate, using a standard graphics board can considerably speed up the overall performance of tasks such as computing Fourier holograms in real time.

GPU-Based Realistic Fur Rendering

Based on the multi-layer textures approach, a new method of rendering real-time fur is proposed. The method takes full advantage of the new features of current Graphic Processing Units to compute the lighting of fur precisely and fast. More importantly, the self-shadow effect of fur and the soft shadow phenomenon over fur are simulated successfully, which enhances the rendering effects greatly. Experiments show that the method can deal with moderate models in real-time, which is very valuable for the applications of fur in many fields, such as cinema, video game and virtual reality etc.