Understanding how dynamic patterns vary across large spatiotemporal ensembles is essential in many scientific domains. In fluid dynamics, for instance, researchers analyze how splash patterns in droplet impact experiments change with physical parameters such as fluid type or impact velocity. These experiments produce large volumes of data where patterns differ in size, shape, and duration, making manual analysis tedious and error-prone. Recently, interactive visualization approaches have been developed to assist analysis using learned similarity models for pattern-based querying. However, they assume fixed-size inputs and only support single-pattern queries, thus limiting their effectiveness for multiscale, multi-pattern analysis and exploration of ensembles. In this paper, we present a visual analysis approach for the interactive exploration of spatiotemporal ensembles through multiscale pattern querying. Our approach extends an existing similarity model to support variable-sized patterns, allowing users to define queries by selecting examples directly on visualized data. Coordinated views enable interactive querying, comparison, and analysis of pattern occurrences and relate pattern occurrences to ensemble parameters. A guidance mechanism supports the user in finding underexplored regions. We demonstrate the utility of our approach on synthetic and real-world datasets. Domain expert feedback confirms that the approach is intuitive, easy to use, and effective for revealing parameter-pattern relationships.

Foveated rendering methods usually reduce spatial resolution in the periphery of the users' view. However, using foveated rendering to reduce temporal resolution, i.e., rendering frame rate, seems less explored. In this work, we present the results of a user study investigating the perceptual effects of foveated temporal resolution reduction, where only the temporal resolution (frame rate) is reduced in the periphery without affecting spatial quality (pixel density). In particular, we investigated the perception of temporal resolution artifacts caused by reducing the frame rate dependent on the eccentricity of the user's gaze. Our user study with 15 participants was conducted in a virtual reality setting using a head-mounted display. Our results indicate that it was possible to reduce average rendering costs, i.e., the number of rendered pixels, to a large degree before participants consistently reported perceiving temporal artifacts.

Abstract The explosion of data in astronomy has resulted in an era of unprecedented opportunities for discovery. The GAIA mission's catalog, containing a large number of light sources (mostly stars) with several parameters such as sky position and proper motion, is playing a significant role in advancing astronomy research and has been crucial in various scientific breakthroughs over the past decade. In its current release, more than 200 million stars contain a calibrated continuous spectrum, which is essential for characterizing astronomical information such as effective temperature and surface gravity, and enabling complex tasks like interstellar extinction detection and narrow‐band filtering. Even though numerous studies have been conducted to visualize and analyze the data in the SciVis and AstroVis communities, no work has attempted to leverage spectral information for visualization in real‐time. Interactive exploration of such complex, massive data presents several challenges for visualization. This paper introduces a novel multi‐resolution, spectrum‐preserving data structure and a progressive, real‐time visualization algorithm to handle the sheer volume of the data efficiently, enabling interactive visualization and exploration of the whole catalog's spectra. We show the efficiency of our method with our open‐source, interactive, web‐based tool for exploring the GAIA catalog, and discuss astronomically relevant use cases of our system.

Abstract Varying the input parameters of simulations or experiments often leads to different classes of results. Parameter sensitivity analysis in this context includes estimating the sensitivity to the individual parameters, that is, to understand which parameters contribute most to changes in output classifications and for which parameter ranges these occur. We propose a novel visual parameter sensitivity analysis approach based on Voronoi cell interfaces between the sample points in the parameter space to tackle the problem. The Voronoi diagram of the sample points in the parameter space is first calculated. We then extract Voronoi cell interfaces which we use to quantify the sensitivity to parameters, considering the class label information of each sample's corresponding output. Multiple visual encodings are then utilized to represent the cell interface transitions and class label distribution, including stacked graphs for local parameter sensitivity. We evaluate the approach's expressiveness and usefulness with case studies for synthetic and real‐world datasets.

We developed a new approach comprised of different visualizations for the comparative spatio-temporal analysis of displacement processes in porous media. We aim to analyze and compare ensemble datasets from experiments to gain insight into the influence of different parameters on fluid flow. To capture the displacement of a defending fluid by an invading fluid, we first condense an input image series to a single time map. From this map, we generate a spatio-temporal flow graph covering the whole process. This graph is further simplified to only reflect topological changes in the movement of the invading fluid. Our interactive tools allow the visual analysis of these processes by visualizing the graph structure and the context of the experimental setup, as well as by providing charts for multiple metrics. We apply our approach to analyze and compare ensemble datasets jointly with domain experts, where we vary either fluid properties or the solid structure of the porous medium. We finally report the generated insights from the domain experts and discuss our contribution's advantages, generality, and limitations.

In visualization research, reproducibility is often not taken into account as a design goal. Reproducibility is desirable beyond ensuring scientific rigor: In the long-term, it helps advance sustainable visualization research, and in the short-term, it supports domain experts using the visualizations in their daily work. Designing for and ensuring reproducibility introduces costs, such as as storage cost or infrastructure maintenance. We propose a typology of reproducibility aspects that can be considered during visualization design. We then propose practical strategies to improve reproducibility and discuss their initial and running costs, their limits, and their tradeoffs. Finally, we discuss three concrete visualization approaches in the context of our typology and strategies.

We present an efficient raycasting algorithm for rendering Volumetric Depth Images (VDIs), and we show how it can be used in a remote visualization setting with VDIs generated and streamed from a remote server. VDIs are compact view-dependent volume representations that enable interactive visualization of large volumes at high frame rates by decoupling viewpoint changes from expensive rendering calculations. However, current rendering approaches for VDIs struggle with achieving interactive frame rates at high image resolutions. Here, we exploit the properties of perspective projection to simplify intersections of rays with the view-dependent frustums in a VDI and leverage spatial smoothness in the volume data to minimize memory accesses. Benchmarks show that responsive frame rates can be achieved close to the viewpoint of generation for HD display resolutions, providing high-fidelity approximate renderings of Gigabyte-sized volumes. We also propose a method to subsample the VDI for preview rendering, maintaining high frame rates even for large viewpoint deviations. We provide our implementation as an extension of an established open-source visualization library.

Data visualization relies on efficient rendering to allow users to interactively explore and understand their data. However, achieving interactive frame rates is often challenging, especially for high-resolution displays or large datasets. In computer graphics, several methods temporally reconstruct full-resolution images from multiple consecutive lower-resolution frames. Besides providing temporal image stability, they amortize the rendering costs over multiple frames and thus improve the minimum frame rate. We present a method that adopts this idea to accelerate 2D information visualization, without requiring any changes to the rendering itself. By exploiting properties of orthographic projection, our method significantly improves rendering performance while minimizing the loss of image quality during camera manipulation. For static scenes, it quickly converges to the full-resolution image. We discuss the characteristics of our method concerning rendering performance and image quality and the corresponding trade-offs. Finally, we present extensive rendering benchmarks to examine real-world performance for examples of parallel coordinates and scatterplot matrix visualizations, and discuss appropriate application scenarios and contraindications for usage.

Abstract We developed a new visualization approach to gain a better understanding of the displacement of one fluid phase by another in porous media. This is based on a recent experimental parameter study with varying capillary numbers and viscosity ratios. We analyse the temporal evolution of characteristic values in this two‐phase flow scenario and discuss how to directly compare experiments across different temporal scales. To enable spatio‐temporal analysis, we introduce a new abstract visual representation showing which paths through the porous medium were occupied and for how long. These transport networks allow to assess the impact of different acting forces and they are designed to yield expressive comparability and linking to the experimental parameter space both supported by additional visual cues. This joint work of porous media experts and visualization researchers yields new insights regarding two‐phase flow on the microscale, and our visualization approach contributes towards the overarching goal of the domain scientists to characterize porous media flow based on capillary numbers and viscosity ratios.

Achieving high rendering quality in the visualization of large particle data, for example from large-scale molecular dynamics simulations, requires a significant amount of sub-pixel super-sampling, due to very high numbers of particles per pixel. Although it is impossible to super-sample all particles of large-scale data at interactive rates, efficient occlusion culling can decouple the overall data size from a high effective sampling rate of visible particles. However, while the latter is essential for domain scientists to be able to see important data features, performing occlusion culling by sampling or sorting the data is usually slow or error-prone due to visibility estimates of insufficient quality. We present a novel probabilistic culling architecture for super-sampled high-quality rendering of large particle data. Occlusion is dynamically determined at the sub-pixel level, without explicit visibility sorting or data simplification. We introduce confidence maps to probabilistically estimate confidence in the visibility data gathered so far. This enables progressive, confidence-based culling, helping to avoid wrong visibility decisions. In this way, we determine particle visibility with high accuracy, although only a small part of the data set is sampled. This enables extensive super-sampling of (partially) visible particles for high rendering quality, at a fraction of the cost of sampling all particles. For real-time performance with millions of particles, we exploit novel features of recent GPU architectures to group particles into two hierarchy levels, combining fine-grained culling with high frame rates.

We introduce inverted stippling , a method to mimic an inversion technique used by artists when performing stippling. To this end, we extend Linde-Buzo-Gray (LBG) stippling to multi-class LBG (MLBG) stippling with multiple layers. MLBG stippling couples the layers stochastically to optimize for per-layer and overall blue-noise properties. We propose a stipple-based filling method to generate solid color backgrounds for inverting areas. Our experiments demonstrate the effectiveness of MLBG in terms of reducing overlapping and intensity accuracy. In addition, we showcase MLBG with color stippling and dynamic multi-class blue-noise sampling, which is possible due to its support for temporal coherence.

Collaborative exploration of scientific data sets across large high-resolution displays requires both high visual detail as well as low-latency transfer of image data (oftentimes inducing the need to trade one for the other). In this work, we present a system that dynamically adapts the encoding quality in such systems in a way that reduces the required bandwidth without impacting the details perceived by one or more observers. Humans perceive sharp, colourful details, in the small foveal region around the centre of the field of view, while information in the periphery is perceived blurred and colourless. We account for this by tracking the gaze of observers, and respectively adapting the quality parameter of each macroblock used by the H.264 encoder, considering the so-called visual acuity fall-off. This allows to substantially reduce the required bandwidth with barely noticeable changes in visual quality, which is crucial for collaborative analysis across display walls at different locations. We demonstrate the reduced overall required bandwidth and the high quality inside the foveated regions using particle rendering and parallel coordinates.

Two-dimensional height fields are the most common data structure used for storing and rendering of terrain in offline rendering and especially real-time computer graphics. By its very nature, a height field cannot store terrain structures with multiple vertical layers such as overhanging cliffs, caves, or arches. This restriction does not apply to volumetric data structures. However, the workflow of manual modelling and editing of volumetric terrain usually is tedious and very time-consuming. Therefore, we propose to use three-dimensional curve-based primitives to efficiently model prominent, large-scale terrain features. We present a technique for volumetric generation of a complete terrain surface from the sparse input data by means of diffusion-based algorithms. By combining an efficient, feature-based toolset with a volumetric terrain representation, the modelling workflow is accelerated and simplified while retaining the full artistic freedom of volumetric terrains. Feature Curves also contain material information that can be complemented with local details by using per-face texture mapping. All stages of our method are GPU-accelerated using compute shaders to ensure interactive editing of terrain. Please note that this paper is an extended version of our previously published work [1] .

Molecular dynamics (MD) simulations are crucial to investigating important processes in physics and thermodynamics. The simulated atoms are usually visualized as hard spheres with Phong shading, where individual particles and their local density can be perceived well in close-up views. However, for large-scale simulations with 10 million particles or more, the visualization of large fields-of-view usually suffers from strong aliasing artifacts, because the mismatch between data size and output resolution leads to severe under-sampling of the geometry. Excessive super-sampling can alleviate this problem, but is prohibitively expensive. This paper presents a novel visualization method for large-scale particle data that addresses aliasing while enabling interactive high-quality rendering. We introduce the novel concept of screen-space normal distribution functions (S-NDFs) for particle data. S-NDFs represent the distribution of surface normals that map to a given pixel in screen space, which enables high-quality re-lighting without re-rendering particles. In order to facilitate interactive zooming, we cache S-NDFs in a screen-space mipmap (S-MIP). Together, these two concepts enable interactive, scale-consistent re-lighting and shading changes, as well as zooming, without having to re-sample the particle data. We show how our method facilitates the interactive exploration of real-world large-scale MD simulation data in different scenarios.

Particle data are commonly visualized by rendering a sphere for each particle. Since interactive rendering usually relies on fast local lighting, the spatial arrangement of the spheres is often very hard to perceive. That is, larger functional structures formed by the particles are not easily recognizable. Using global effects such as ambient occlusion or shadows adds important depth cues. In this work, we present Implicit Sphere Shadow Maps (ISSM), an application-tailored approach for large, dynamic particle data sets. This approach can be combined with state-of-the-art object-space ambient occlusion to further emphasize the spatial structure of molecules. We compare our technique against state-of-the-art methods for interactive rendering with respect to image quality and performance.

Two-dimensional heightfields are the most common data structure used for storing and rendering of terrain in offline rendering and especially real-time computer graphics. By its very nature, a heightfield cannot store terrain structures with multiple vertical layers such as overhanging cliffs, caves, or arches. This restriction does not apply to volumetric data structures. However, the workflow of manual modelling and editing of volumetric terrain usually is tedious and very time-consuming. Therefore, we propose to use three-dimensional curve-based primitives to efficiently model prominent, large-scale terrain features. We present a technique for volumetric generation of a complete terrain surface from the sparse input data by means of diffusion-based algorithms. By combining an efficient, feature-based toolset with a volumetric terrain representation, the modelling workflow is accelerated and simplified while retaining the full artistic freedom of volumetric terrains. All stages of our method are GPU-accelerated using compute shaders to ensure interactive editing of terrain.

Analysis of software performance typically takes into account clock cycles and memory consumption at each sampling point in time. Although this is a valid strategy, we argue that it is also worth investigating data and control flow structures, as observed using memory traces and call stacks, because of their importance for performance engineering. In this work, we present a visual approach to memory profiling that supports analysis of memory layout, access patterns, and aliasing in correlation to program execution. Our method leverages language-agnostic dynamic code instrumentation to minimize the impact of tracing on performance, i.e., the application remains usable on commodity hardware. The profiled data is then clustered and visualized using a density-based scatter plot. If debug symbols are available, the scatter plot is augmented by a flame graph to ease linking to the high-level source code. Our visualization helps software engineers to identify runtime behavior by relating memory addresses to instruction execution. We demonstrate our approach using a set of examples revealing different memory access patterns and discuss their influence on software performance.

We present Molecular Surface Maps, a novel, view-independent, and concise representation for molecular surfaces. It transfers the well-known world map metaphor to molecular visualization. Our application maps the complex molecular surface to a simple 2D representation through a spherical intermediate, the Molecular Surface Globe. The Molecular Surface Map concisely shows arbitrary attributes of the original molecular surface, such as biochemical properties or geometrical features. This results in an intuitive overview, which allows researchers to assess all molecular surface attributes at a glance. Our representation can be used as a visual summarization of a molecule's interface with its environment. In particular, Molecular Surface Maps simplify the analysis and comparison of different data sets or points in time. Furthermore, the map representation can be used in a Space-time Cube to analyze time-dependent data from molecular simulations without the need for animation. We show the feasibility of Molecular Surface Maps for different typical analysis tasks of biomolecular data.

Abstract In this report, we review the current state of the art of web‐based visualization applications. Recently, an increasing number of web‐based visualization applications have emerged. This is due to the fact that new technologies offered by modern browsers greatly increased the capabilities for visualizations on the web. We first review these technical aspects that are enabling this development. This includes not only improvements for local rendering like WebGL and HTML5, but also infrastructures like grid or cloud computing platforms. Another important factor is the transfer of data between the server and the client. Therefore, we also discuss advances in this field, for example methods to reduce bandwidth requirements like compression and other optimizations such as progressive rendering and streaming. After establishing these technical foundations, we review existing web‐based visualization applications and prototypes from various application domains. Furthermore, we propose a classification of these web‐based applications based on the technologies and algorithms they employ. Finally, we also discuss promising application areas that would benefit from web‐based visualization and assess their feasibility based on the existing approaches.

Processing and visualizing large scale volumetric and geometric datasets is mission critical in an increasing number of applications in academic research as well as in commercial enterprise. Often the datasets are, or can be processed to become, sparse. In this paper, we present VoxLink, a novel approach to render sparse volume data in a memory-efficient manner enabling interactive rendering on common, offthe- shelf graphics hardware. Our approach utilizes current GPU architectures for voxelizing, storing, and visualizing such datasets. It is based on the idea of perpixel linked lists (ppLL), an A-buffer implementation for order-independent transparency rendering. The method supports voxelization and rendering of dense semi-transparent geometry, sparse volume data, and implicit surface representations with a unified data structure. The proposed data structure also enables efficient simulation of global lighting effects such as reflection, refraction, and shadow ray evaluation.

In many disciplines, such as computer aided drug design, multiple simulation runs are performed with varying parameters, yielding ensembles of data sets. Comparative visualisation of these simulation results can help understanding the influence of different parameters. However, researchers might need to compare large numbers of variants. Single desktop monitors often do not have the resolution and screen size required for showing a whole ensemble at once with sufficient detail. Wall-sized high-resolution displays can be a solution for this problem. Although a number of studies has been conducted on how large high-resolution displays affect the speed and accuracy of certain tasks, only few of them are related to actual scientific visualisation tasks. We built a system for comparative visualisation of simulation results that can be used with conventional desktop monitors and with large high-resolution displays. We conducted a study using biochemical simulation data to evaluate the impact of screen size and 3D stereo display on a comparison task.

The term code ownership is used in software engineering to describe who authored a certain piece of software. Code ownership is commonly determined by investigating the data from version control systems, defining the owner as the person who contributed the most lines to a file, module, etc. Existing visualisation for ownership usually relies on per-line annotations from the version control system, thus only conveying the information who changed a line the last time, potentially adding some visual cue about how old the respective change is. This, however, can be misleading, because any change of even a single character changes ownership. In this paper, we propose a visualisation that accounts for fractional ownership changes over time on a per-character basis. Our technique incorporates visual cues to convey that typical definitions of ownership have an inherent uncertainty and provides details on the cause of this uncertainty on demand. For our definition of code ownership being a low-level one, we propose implementing the visualisation as an in-situ visualisation in the code editor of modern development environments. We will show examples of the efficacy of our approach and discuss its advantages and disadvantages compared to conventional line-based ownership.

Graph visualization is essential for the analysis of networks and relational data sets. Often, most of the effort is expended on computing sophisticated layouts of the visual representation of the graph. Even though this is increasingly accelerated by use of graphics processing units (GPUs), the rendering is often considered as circumstantial. In this paper, we present a coherent approach to graph visualization that utilizes all features of modern GPUs. We describe specialized data structures and our GPU-centric pipeline for computing and rendering a layout, while enabling steering and interaction. We evaluate technical aspects of our approach as well as its applicability to huge real-world graphs.

Visualization applications nowadays not only face increasingly larger datasets, but have to solve increasingly complex research questions. They often require more than a single algorithm and consequently a software solution will exceed the possibilities of simple research prototypes. Well-established systems intended for such complex visual analysis purposes have usually been designed for classical, mesh-based graphics approaches. For particle-based data, however, existing visualization frameworks are too generic - e.g. lacking possibilities for consistent low-level GPU optimization for high-performance graphics - and at the same time are too limited - e.g. by enforcing the use of structures suboptimal for some computations. Thus, we developed the system softwareMegaMol for visualization research on particle-based data. On the one hand, flexible data structures and functional module design allow for easy adaption to changing research questions, e.g. studying vapors in thermodynamics, solid material in physics, or complex functional macromolecules like proteins in biochemistry. Therefore, MegaMol is designed as a development framework. On the other hand, common functionality for data handling and advanced rendering implementations are available and beneficial for all applications. We present several case studies of work implemented using our system as well as a comparison to other freely available or open source systems.

We present a visual analysis application for protein simulations that allows researchers to investigate dynamic data interactively. Special focus lies on the analysis of cavities and binding sites, which play a critical role in protein function. Cavities are extracted and classified. The surface area of all cavities and the diameter of pockets and channels is computed to provide evidence for their accessibility. These values are also plotted in 2D graphs for a quantitative time-dependent analysis. For dynamic simulation data sets, the cavities are tracked to show their stability over time. The user is provided with a range of application-related parameters to interactively adjust the analysis algorithms. A sequence diagram shows the structure of the protein and additional annotations like binding sites. Furthermore, all views support brushing and linking for consistent selection and filtering. All algorithmic steps are implemented to run interactively on a commodity workstation. As a result, the user can immediately see the effect of a parameter change. This enables the real-time analysis of a running simulation for in-situ visualization.

As systems grow at an increasing pace to keep up with the demand for larger data and ever more detailed simulations, failures also occur more frequently on various levels. Traditionally, fault tolerance for visualization is either disregarded due to performance reasons or handled analogously to the state of the art for high performance computing (HPC), for example rollback-recovery. In this paper we summarize the inherent challenges of distributed visualization as well as the applicability and differences of fault tolerance methods used in HPC and distributed visualization. We derive complementary methods to increase the resilience of such systems and present visualization specific methods for failure semantics and masking.

Abstract The comparison of molecular surface attributes is of interest for computer aided drug design and the analysis of biochemical simulations. Due to the non‐rigid nature of molecular surfaces, partial shape matching is feasible for mapping two surfaces onto each other. We present a novel technique to obtain a mapping relation between two surfaces using a deformable model approach. This relation is used for pair‐wise comparison of local surface attributes (e.g. electrostatic potential). We combine the difference value as well as the comparability as derived from the local matching quality in a 3D molecular visualization by mapping them to color. A 2D matrix shows the global dissimilarity in an overview of different data sets in an ensemble. We apply our visualizations to simulation results provided by collaborators from the field of biochemistry to evaluate the effectiveness of our results.

Abstract We present a coordinated‐view application for the analysis of molecular surface features like cavities, channels and pockets. Our tool employs object‐space ambient occlusion for the detection of such features and tracks them over time. It offers time‐dependent graphs of metrics concerning those features and allows analyzing the temporal relationship of the features, i.e. when they (dis)appear, split or merge and which features participate in each of these events. The automated analysis process is performed in real time while the user interactively explores a dynamic data set. The system supports linking and brushing to allow for a user‐guided visual analysis based on different aspects of the data. We demonstrate the effectiveness of our approach by applying it to data sets from biochemistry and report the insights that can be gained. We also evaluate the benefits of our method with respect to recent advancements in the field. The algorithmic pipeline leverages the computing power of modern GPUs, thus achieving interactive frame rates without any precomputation for fully dynamic data sets.

The global scheduler of a current GPU distributes thread blocks to symmetric multiprocessors (SM), which schedule threads for execution with the granularity of a warp. Threads in a warp execute the same code path in lockstep, which potentially leads to a large amount of wasted cycles for divergent control flow. In order to overcome this general issue of SIMT architectures, we propose techniques to relax divergence on the fly within a computation kernel in order to achieve a much higher total utilization of processing cores. We propose techniques for branch and loop divergence (which may also be combined) switching to suitable tasks during a GPU kernel run every time divergence occurs. Our newly introduced techniques can easily be applied to arbitrary iterative algorithms and we evaluate the performance and effectiveness of our approach exemplarily via synthetic and real world applications.

Metal oxides are important for many technical applications. For example alumina (aluminum oxide) is the most commonly-used ceramic in microelectronic devices thanks to its excellent properties. Experimental studies of these materials are increasingly supplemented with computer simulations. Molecular dynamics (MD) simulations can reproduce the material behavior very well and are now reaching time scales relevant for interesting processes like crack propagation. In this work we focus on the visualization of induced electric dipole moments on oxygen atoms in crack propagation simulations. The straightforward visualization using glyphs for the individual atoms, simple shapes like spheres or arrows, is insufficient for providing information about the data set as a whole. As our contribution we show for the first time that fractional anisotropy values computed from the local neighborhood of individual atoms of MD simulation data depict important information about relevant properties of the field of induced electric dipole moments. Iso surfaces in the field of fractional anisotropy as well as adjustments of the glyph representation allow the user to identify regions of correlated orientation. We present novel and relevant findings for the application domain resulting from these visualizations, like the influence of mechanical forces on the electrostatic properties.

Abstract Molecular dynamics simulations are a principal tool for studying molecular systems. Such simulations are used to investigate molecular structure, dynamics, and thermodynamical properties, as well as a replacement for, or complement to, costly and dangerous experiments. With the increasing availability of computational power the resulting data sets are becoming increasingly larger, and benchmarks indicate that the interactive visualization on desktop computers poses a challenge when rendering substantially more than millions of glyphs. Trading visual quality for rendering performance is a common approach when interactivity has to be guaranteed. In this paper we address both problems and present a method for high‐quality visualization of massive molecular dynamics data sets. We employ several optimization strategies on different levels of granularity, such as data quantization, data caching in video memory, and a two‐level occlusion culling strategy: coarse culling via hardware occlusion queries and a vertex‐level culling using maximum depth mipmaps. To ensure optimal image quality we employ GPU raycasting and deferred shading with smooth normal vector generation. We demonstrate that our method allows us to interactively render data sets containing tens of millions of high‐quality glyphs.

Particle-based simulations are a popular tool for researchers in various sciences. In combination with the availability of ever larger COTS clusters and the consequently increasing number of simulated particles the resulting datasets pose a challenge for real-time visualization. Additionally the semantic density of the particles exceeds the possibilities of basic glyphs, like splats or spheres and results in dataset sizes larger by at least an order of magnitude. Interactive visualization on common workstations requires a careful optimization of the data management, especially of the transfer between CPU and GPU. We propose a flexible benchmarking tool along with a series of tests to allow the evaluation of the performance of different CPU/GPU combinations in relation to a particular implementation. We evaluate different uploading strategies and rendering methods for point-based compound glyphs suitable for representing the aforementioned datasets. CPU and GPU-based approaches are compared with respect to their rendering and storage efficiency to point out the optimal solution when dealing with time-dependent datasets. The results of our research are of general interest since they can be transferred to other applications where CPU-GPU bandwidth and a high number of graphical primitives per dataset pose a problem. The employed tool set for streamlining the measurement process is made publicly available.

A current research topic in molecular thermodynamics is the condensation of vapor to liquid and the investigation of this process at the molecular level. Condensation is found in many physical phenomena, e.g. the formation of atmospheric clouds or the processes inside steam turbines, where a detailed knowledge of the dynamics of condensation processes will help to optimize energy efficiency and avoid problems with droplets of macroscopic size. The key properties of these processes are the nucleation rate and the critical cluster size. For the calculation of these properties it is essential to make use of a meaningful definition of molecular clusters, which currently is a not completely resolved issue. In this paper a framework capable of interactively visualizing molecular datasets of such nucleation simulations is presented, with an emphasis on the detected molecular clusters. To check the quality of the results of the cluster detection, our framework introduces the concept of flow groups to highlight potential cluster evolution over time which is not detected by the employed algorithm. To confirm the findings of the visual analysis, we coupled the rendering view with a schematic view of the clusters' evolution. This allows to rapidly assess the quality of the molecular cluster detection algorithm and to identify locations in the simulation data in space as well as in time where the cluster detection fails. Thus, thermodynamics researchers can eliminate weaknesses in their cluster detection algorithms. Several examples for the effective and efficient usage of our tool are presented.

We propose a new approach for the visualization of hyperstreamlines, which offers potential for better scalability than the conventional polygon-based approach. Our method circumvents the bandwidth bottleneck between the CPU and GPU by transmitting a small set of parameters for each tube segment and generates the surface directly on the GPU using the classical sphere tracing approach. This reduces the load on the CPU that would otherwise need to provide a suitable level-of-detail representation of the scene, while offering even higher quality in the resulting surfaces since every fragment is traced individually. We demonstrate the effectiveness of this approach by comparing it to the performance and output of conventional visualization tools in the application area of diffusion tensor imaging of human brain MR scans. The method presented here can also be utilized to generate other types of surfaces on the GPU that are too complex to handle with direct ray casting and can therefore be adapted for other applications.

We present a novel visualization method for mono- and dipolar molecular simulations from thermodynamics that takes advantage of modern graphics hardware to interactively render specifically tailored glyphs. Our approach allows domain experts to visualize the results of molecular dynamics simulations with a higher number of particles than before and furthermore offers much better visual quality. We achieve this by transferring only visualization parameters to the GPU and by generating implicit surfaces directly in the fragment program. As a result, we can render up to 500.000 glyphs with about 10 fps displaying all the simulation results as geometrical properties that resemble the classical abstract representation used in this research area. Thus we enable researchers to visually assess the results of simulations of greater scale than before. We believe that the proposed method can be generalized to create other kinds of parametrized surfaces directly on graphics hardware to overcome the bandwidth bottleneck that exists between CPU and GPU.