EGPGV15: Eurographics Symposium on Parallel Graphics and Visualization

Permanent URI for this collection

https://diglib.eg.org/handle/10.2312/12465

Browse

Now showing 1 - 13 of 13

Visualizing Groundwater Flow Through Karst Limestone
(The Eurographics Association, 2015) Knoll, Aaron; Brownlee, Carson; Navrátil, Paul; Cunningham, Kevin J.; Sukop, Michael C.; Garcia, Sadé; C. Dachsbacher and P. Navrátil
Water management is critical in Florida where freshwater is often rare or, in times of flooding, overabundant and seawater frequently contaminates available sources. Professor Michael Sukop from Florida International University, his student, Sadé Garcia, and Dr. Kevin Cunningham of the United States Geological Survey are developing techniques to better understand flow through aquifers in South Florida, which are vital sources of freshwater. 3D flow simulations of groundwater through Computed Tomography (CT) data from samples of karst limestone allowed them to more accurately predict the permeability values of the rock in their tests than existing laboratory measurement techniques. Researchers at TACC visualized these simulations by developing a rendering library which can render photo-realistic images using a path tracer built with Intel's Embree ray tracing kernels by intercepting calls to the OpenGL API. Using this software, they were able to generate significant improvements over native OpenGL rendering in existing tools and better illustrate the flow through thumb-sized holes in the limestone.
Large-Scale Parallel Visualization of Particle-Based Simulations using Point Sprites and Level-Of-Detail
(The Eurographics Association, 2015) Rizzi, Silvio; Hereld, Mark; Insley, Joseph; Papka, Michael E.; Uram, Thomas; Vishwanath, Venkatram; C. Dachsbacher and P. Navrátil
Recent large-scale particle-based simulations are generating vast amounts of data posing a challenge to visualization algorithms. One possibility for addressing this challenge is to map particles into a regular grid for volume rendering, which carries the disadvantages of inefficient use of memory and undesired losses of dynamic range. As an alternative, we propose a method to efficiently visualize these massive particle datasets using point rendering techniques with neither loss of dynamic range nor memory overheads. In addition, a hierarchical reorganization of the data is desired to deliver meaningful visual representations of a large number of particles in a limited number of pixels, preserving point locality and also helping achieve interactive frame rates. In this paper, we present a framework for parallel rendering of large-scale particle data sets combining point sprites and z-ordering. The latter is used to create a multi level representation of the data which helps improving frame rates. Performance and scalability are evaluated on a GPU-based visualization cluster, scaling up to 128 GPUs. Results using particle datasets of up to 32 billion particles are shown.
Contour Tree Depth Images For Large Data Visualization
(The Eurographics Association, 2015) Biedert, Tim; Garth, Christoph; C. Dachsbacher and P. Navrátil
High-fidelity simulation models on large-scale parallel computer systems can produce data at high computational throughput, but modern architectural trade-offs make full persistent storage to the slow I/O subsystem prohibitively costly with respect to time. We demonstrate the feasibility and potential of combining in situ topological contour tree analysis and compact image-based data representation to address this problem. Our experiments show significant reductions in storage requirements using topology-guided layered depth imaging, while preserving flexibility for explorative visualization and analysis. Our approach represents an effective and easy-to-control trade-off between storage overhead and visualization fidelity for large data visualization.
Packet-Oriented Streamline Tracing on Modern SIMD Architectures
(The Eurographics Association, 2015) Hentschel, Bernd; Göbbert, Jens Henrik; Klemm, Michael; Springer, Paul; Schnorr, Andrea; Kuhlen, Torsten W.; C. Dachsbacher and P. Navrátil
The advection of integral lines is an important computational kernel in vector field visualization. We investigate how this kernel can profit from vector (SIMD) extensions in modern CPUs. As a baseline, we formulate a streamline tracing algorithm that facilitates auto-vectorization by an optimizing compiler. We analyze this algorithm and propose two different optimizations. Our results show that particle tracing does not per se benefit from SIMD computation. Based on a careful analysis of the auto-vectorized code, we propose an optimized data access routine and a re-packing scheme which increases average SIMD efficiency. We evaluate our approach on three different, turbulent flow fields. Our optimized approaches increase integration performance up to 5:6 over our baseline measurement. We conclude with a discussion of current limitations and aspects for future work.
Out-of-Core Framework for QEM-based Mesh Simplification
(The Eurographics Association, 2015) Ozaki, Hiromu; Kyota, Fumihito; Kanai, Takashi; C. Dachsbacher and P. Navrátil
In mesh simplification, in-core based methods using Quadric Error Metric (QEM), which apply a sequence of edge-collapse operations, can generate high-quality meshes while preserving shape features. However, these methods cannot be applied to huge meshes with more than 100 million faces, because they require considerable memory. On the other hand, the quality of simplified meshes by previous out-of-core algorithms tends to be insufficient. In this paper, we propose an out-of-core framework to establish high-quality QEM-based simplification for huge meshes. To simplify a huge mesh using limited memory, the mesh is first partitioned into a set of patches in the out-of core framework using linear classifiers which are trained by clustered points based on the machine learning approach. Also, a scheme to guarantee the exact matching of boundary vertices between neighbor patches is proposed even when each patch is simplified independently. Based on this scheme, out-of-core simplification is established while generating a simplified mesh with almost the same quality as that of the in-core QEM-based method. We apply the proposed method to multiple models including huge meshes and show the superiority of our method over previous state-of-the-art methods in terms of the quality of simplified meshes.
SIMD Parallel Ray Tracing of Homogeneous Polyhedral Grids
(The Eurographics Association, 2015) Rathke, Brad; Wald, Ingo; Chiu, Kenneth; Brownlee, Carson; C. Dachsbacher and P. Navrátil
Efficient visualization of unstructured data is vital for domain scientists, yet is often impeded by techniques which rely on intermediate representations that consume time and memory, require resampling data, or inefficient implementations of direct ray tracing methods. Previous work to accelerate rendering of unstructured grids have focused on the use of GPUs that are not available in many large-scale computing systems. In this paper, we present a new technique for directly visualizing unstructured grids using a software ray tracer built as a module for the OSPRay ray tracing framework from Intel. Our method is capable of implicit isosurface rendering and direct volume ray casting homogeneous grids of hexahedra, tetrahedra, and multi-level datasets at interactive frame rates on compute nodes without a GPU using an open-source, production-level ray tracing framework that scales with variable SIMD widths and thread counts.
Frontmatter: EG PGV 2015 - Eurographics Symposium on Parallel Graphics and Visualization
(Eurographics Association, 2015) Dachsbacher, Carsten; Navrátil, Paul; -
Visualization of High-Resolution Weather Model Data
(The Eurographics Association, 2015) Liu, Si; Foss, Greg; Abram, Greg; Bowen, Anne; C. Dachsbacher and P. Navrátil
Significant bottlenecks from I/O required novel techniques be used for a high-resolution severe weather simulation and visualization in the region centered around O'Hare International Airport. The Raytheon Company collaborated with the Texas Advanced Computing Center (TACC) and National Center for Atmospheric Research (NCAR) to utilize the Weather Research and Forecasting Model (WRF). WRF is a state-of-the-art parallel mesoscale numerical weather prediction system. In order to minimize the data read from disk, we develop a custom importer in Paraview to read the WRF output data directly from HDF5, de-stagger variables as required, convert the grid coordinates to first create accurate real-world (longitude, latitude, elevation) coordinates and then project them into a WGS84 coordinates. This projection enabled us to correctly co-locate the WRF data with an aviation map base layer that was texture mapped onto a refined grid to approximate the non-linearity of a (longitude, latitude) coordinate system.
Memory-Efficient On-The-Fly Voxelization of Particle Data
(The Eurographics Association, 2015) Zirr, Tobias; Dachsbacher, Carsten; C. Dachsbacher and P. Navrátil
In this paper we present a novel GPU-friendly real-time voxelization technique for rendering homogeneous media that is defined by particles, e.g. fluids obtained from particle-based simulations such as Smoothed Particle Hydrodynamics (SPH). Our method computes view-adaptive binary voxelizations with on-the-fly compression of a tiled perspective voxel grid, achieving higher resolutions than previous approaches. It allows for interactive generation of realistic images, enabling advanced rendering techniques such as ray casting-based refraction and reflection, light scattering and absorption, and ambient occlusion. In contrast to previous methods, it does not rely on preprocessing such as expensive, and often coarse, scalar field conversion or mesh generation steps. Our method directly takes unsorted particle data as input. It can be further accelerated by identifying fully populated simulation cells during simulation. The extracted surface can be filtered to achieve smooth surface appearance.
Volume Rendering Via Data-Parallel Primitives
(The Eurographics Association, 2015) Larsen, Matthew; Labasan, Stephanie; Navrátil, Paul; Meredith, Jeremy; Childs, Hank; C. Dachsbacher and P. Navrátil
Supercomputing designs have recently evolved to include architectures beyond the standard CPU. In response, visualization software must be developed in a manner that obviates the need for porting all visualization algorithms to all architectures. Recent research results indicate that building visualization software on a foundation of dataparallel primitives can meet this goal, providing portability over many architectures, and doing it in a performant way. With this work, we introduce an unstructured data volume rendering algorithm which is composed entirely of data-parallel primitives. We compare the algorithm to community standards, and show that the performance we achieve is similar. That is, although our algorithm is hardware-agnostic, we demonstrate that our performance on GPUs is comparable to code that was written for and optimized for the GPU, and our performance on CPUs is comparable to code written for and optimized for the CPU. The main contribution of this work is in realizing the benefits of data-parallel primitives - portable performance, longevity, and programmability - for volume rendering. A secondary contribution is in providing further evidence of the merits of the data-parallel primitives approach itself.
Visualization of 2DWave Propagation by Huygens' Principle
(The Eurographics Association, 2015) Heßel, Stefan; Fernandes, Oliver; Boblest, Sebastian; Offenhäuser, Philipp; Hoffmann, Malte; Beck, Andrea; Ertl, Thomas; Glass, Colin; Munz, Claus-Dieter; Sadlo, Filip; C. Dachsbacher and P. Navrátil
We present a novel technique to visualize wave propagation in 2D scalar fields. Direct visualization of wave fronts is susceptible to visual clutter and interpretation difficulties due to space-time interference and global influence. To avoid this, we employ Huygens' principle to obtain virtual sources that provide a concise spacetime representation of the overall dynamics by means of elementary waves. We first demonstrate the utility of our overall approach by computing a dense field of virtual sources. This variant offers full insight into space-time wave dynamics in terms of elementary waves, but it reflects the full problem of inverse wave propagation and hence suffers from high costs regarding memory consumption and computation. As an alternative, we therefore provide a less accurate and less generic but more efficient approach. This alternative performs wave front extraction with subsequent Hough transform to identify potential virtual sources. We evaluate both approaches and demonstrate their strengths and weaknesses by means of a GPU-based prototype and an implementation on a Cray XC40 supercomputer, using data from different domains.
TOD-Tree: Task-Overlapped Direct send Tree Image Compositing for Hybrid MPI Parallelism
(The Eurographics Association, 2015) Grosset, A. V. Pascal; Prasad, Manasa; Christensen, Cameron; Knoll, Aaron; Hansen, Charles; C. Dachsbacher and P. Navrátil
Modern supercomputers have very powerful multi-core CPUs. The programming model on these supercomputer is switching from pure MPI to MPI for inter-node communication, and shared memory and threads for intra-node communication. Consequently the bottleneck in most systems is no longer computation but communication between nodes. In this paper, we present a new compositing algorithm for hybrid MPI parallelism that focuses on communication avoidance and overlapping communication with computation at the expense of evenly balancing the workload. The algorithm has three stages: a direct send stage where nodes are arranged in groups and exchange regions of an image, followed by a tree compositing stage and a gather stage. We compare our algorithm with radix-k and binary-swap from the IceT library in a hybrid OpenMP/MPI setting, show strong scaling results and explain how we generally achieve better performance than these two algorithms.
Visualization Showcase: General-Relativistic Black Hole Visualization
(The Eurographics Association, 2015) Müller, Thomas; Boblest, Sebastian; Weiskopf, Daniel; C. Dachsbacher and P. Navrátil
Black holes are among the most fascinating and weird objects in the universe. They distort space and time in their close neighborhood in a way that is far beyond our every day experience. We demonstrate the visual effects of this curved spacetime by means of four-dimensional nonlinear ray tracing applied to an accretion disk around a spinning black hole and a sphere oscillating between two static, charged black holes. We discuss how visualization helps predict and communicate the interesting effects of general relativity, in particular, its geometric effects on light propagation. The nonlinear behavior of light propagation leads to a compute-intensive rendering process; we report on our experiences with highly parallel rendering in this context.

Browse

Browsing EGPGV15: Eurographics Symposium on Parallel Graphics and Visualization by Issue Date

Results Per Page

Sort Options