SCIENTIFIC COMPUTING AND IMAGING INSTITUTE
at the University of Utah

The Helmholtz-Hodge decomposition (HHD) is one of the fundamental theorems of fluids describing the decomposition of a flow field into its divergence-free, curl-free and harmonic components. Solving for an HDD is intimately connected to the choice of boundary conditions which determine the uniqueness and orthogonality of the decomposition. This article points out that one of the boundary conditions used in a recent paper \"Meshless Helmholtz-Hodge decomposition\" [5] is, in general, invalid and provides an analytical example demonstrating the problem. We hope that this clarification on the theory will foster further research in this area and prevent undue problems in applying and extending the original approach.

As scientific applications target exascale, challenges related to data and energy are becoming dominating concerns. For example, coupled simulation workflows are increasingly adopting in-situ data processing and analysis techniques to address costs and overheads due to data movement and I/O. However it is also critical to understand these overheads and associated trade-offs from an energy perspective. The goal of this paper is exploring data-related energy/performance trade-offs for end-to-end simulation workflows running at scale on current high-end computing systems. Specifically, this paper presents: (1) an analysis of the data-related behaviors of a combustion simulation workflow with an in-situ data analytics pipeline, running on the Titan system at ORNL; (2) a power model based on system power and data exchange patterns, which is empirically validated; and (3) the use of the model to characterize the energy behavior of the workflow and to explore energy/performance trade-offs on current as well as emerging systems.

Keywords: SDAV

This paper introduces a novel partition-based regression approach that incorporates topological information. Partition-based regression typically introduce a quality-of-fit-driven decomposition of the domain. The emphasis in this work is on a topologically meaningful segmentation. Thus, the proposed regression approach is based on a segmentation induced by a discrete approximation of the Morse-Smale complex. This yields a segmentation with partitions corresponding to regions of the function with a single minimum and maximum that are often well approximated by a linear model. This approach yields regression models that are amenable to interpretation and have good predictive capacity. Typically, regression estimates are quantified by their geometrical accuracy. For the proposed regression, an important aspect is the quality of the segmentation itself. Thus, this paper introduces a new criterion that measures the topological accuracy of the estimate. The topological accuracy provides a complementary measure to the classical geometrical error measures and is very sensitive to over-fitting. The Morse-Smale regression is compared to state-of-the-art approaches in terms of geometry and topology and yields comparable or improved fits in many cases. Finally, a detailed study on climate-simulation data demonstrates the application of the Morse-Smale regression. Supplementary materials are available online and contain an implementation of the proposed approach in the R package msr, an analysis and simulations on the stability of the Morse-Smale complex approximation and additional tables for the climate-simulation study.

The next generation of methodologies for nuclear reactor Probabilistic Risk Assessment (PRA) explicitly accounts for the time element in modeling the probabilistic system evolution and uses numerical simulation tools to account for possible dependencies between failure events. The Monte-Carlo (MC) and the Dynamic Event Tree (DET) approaches belong to this new class of dynamic PRA methodologies. A challenge of dynamic PRA algorithms is the large amount of data they produce which may be difficult to visualize and analyze in order to extract useful information. We present a software tool that is designed to address these goals. We model a large-scale nuclear simulation dataset as a high-dimensional scalar function defined over a discrete sample of the domain. First, we provide structural analysis of such a function at multiple scales and provide insight into the relationship between the input parameters and the output. Second, we enable exploratory analysis for users, where we help the users to differentiate features from noise through multi-scale analysis on an interactive platform, based on domain knowledge and data characterization. Our analysis is performed by exploiting the topological and geometric properties of the domain, building statistical models based on its topological segmentations and providing interactive visual interfaces to facilitate such explorations. We provide a user's guide to our software tool by highlighting its analysis and visualization capabilities, along with a use case involving data from a nuclear reactor safety simulation.

Keywords: high-dimensional data analysis, computational topology, nuclear reactor safety analysis, visualization

Nuclear simulations are often computationally expensive, time-consuming, and high-dimensional with respect to the number of input parameters. Thus exploring the space of all possible simulation outcomes is infeasible using finite computing resources. During simulation-based probabilistic risk analysis, it is important to discover the relationship between a potentially large number of input parameters and the output of a simulation using as few simulation trials as possible. This is a typical context for performing adaptive sampling where a few observations are obtained from the simulation, a surrogate model is built to represent the simulation space, and new samples are selected based on the model constructed. The surrogate model is then updated based on the simulation results of the sampled points. In this way, we attempt to gain the most information possible with a small number of carefully selected sampled points, limiting the number of expensive trials needed to understand features of the simulation space.

We analyze the specific use case of identifying the limit surface, i.e., the boundaries in the simulation space between system failure and system success. In this study, we explore several techniques for adaptively sampling the parameter space in order to reconstruct the limit surface. We focus on several adaptive sampling schemes. First, we seek to learn a global model of the entire simulation space using prediction models or neighborhood graphs and extract the limit surface as an iso-surface of the global model. Second, we estimate the limit surface by sampling in the neighborhood of the current estimate based on topological segmentations obtained locally.

Our techniques draw inspirations from topological structure known as the Morse-Smale complex. We highlight the advantages and disadvantages of using a global prediction model versus local topological view of the simulation space, comparing several different strategies for adaptive sampling in both contexts. One of the most interesting models we propose attempt to marry the two by obtaining a coarse global representation using prediction models, and a detailed local representation based on topology. Our methods are validated on several analytical test functions as well as a small nuclear simulation dataset modeled after a simplified Pressurized Water Reactor.

Keywords: high-dimensional data analysis, computational topology, nuclear reactor safety analysis, visualization

We investigate the use of a topology-based clustering technique on the data generated by dynamic event tree methodologies. The clustering technique we utilizes focuses on a domain-partitioning algorithm based on topological structures known as the Morse-Smale complex, which partitions the data points into clusters based on their uniform gradient flow behavior. We perform both end state analysis and transient analysis to classify the set of nuclear scenarios. We demonstrate our methodology on a dataset generated for a sodium-cooled fast reactor during an aircraft crash scenario. The simulation tracks the temperature of the reactor as well as the time for a recovery team to fix the passive cooling system. Combined with clustering results obtained previously through mean shift methodology, we present the user with complementary views of the data that help illuminate key features that may be otherwise hidden using a single methodology. By clustering the data, the number of relevant test cases to be selected for further analysis can be drastically reduced by selecting a representative from each cluster. Identifying the similarities of simulations within a cluster can also aid in the drawing of important conclusions with respect to safety analysis.

Understanding and describing expensive black box functions such as physical simulations is a common problem in many application areas. One example is the recent interest in uncertainty quantification with the goal of discovering the relationship between a potentially large number of input parameters and the output of a simulation. Typically, the simulation of interest is expensive to evaluate and thus the sampling of the parameter space is necessarily small. As a result choosing a "good" set of samples at which to evaluate is crucial to glean as much information as possible from the fewest samples. While space-filling sampling designs such as Latin hypercubes provide a good initial cover of the entire domain, more detailed studies typically rely on adaptive sampling: Given an initial set of samples, these techniques construct a surrogate model and use it to evaluate a scoring function which aims to predict the expected gain from evaluating a potential new sample. There exist a large number of different surrogate models as well as different scoring functions each with their own advantages and disadvantages. In this paper we present an extensive comparative study of adaptive sampling using four popular regression models combined with six traditional scoring functions compared against a space-filling design. Furthermore, for a single high-dimensional output function, we introduce a new class of scoring functions based on global topological rather than local geometric information. The new scoring functions are competitive in terms of the root mean squared prediction error but are expected to better recover the global topological structure. Our experiments suggest that the most common point of failure of adaptive sampling schemes are ill-suited regression models. Nevertheless, even given well-fitted surrogate models many scoring functions fail to outperform a space-filling design.

Historically, data creation and storage has always outpaced the infrastructure for its movement and utilization. This trend is increasing now more than ever, with the ever growing size of scientific simulations, increased resolution of sensors, and large mosaic images. Effective exploration of massive scientific models demands the combination of data management, analysis, and visualization techniques, working together in an interactive setting. The ViSUS application framework has been designed as an environment that allows the interactive exploration and analysis of massive scientific models in a cache-oblivious, hardware-agnostic manner, enabling processing and visualization of possibly geographically distributed data using many kinds of devices and platforms.

For general purpose feature segmentation and exploration we discuss a new paradigm based on topological analysis. This approach enables the extraction of summaries of features present in the data through abstract models that are orders of magnitude smaller than the raw data, providing enough information to support general queries and perform a wide range of analyses without access to the original data.

Keywords: Visualization, data analysis, topological data analysis, Parallel I/O

Gigapixel panoramas are an increasingly popular digital image application. They are often created as a mosaic of smaller images composited into a larger single image. The mosaic acquisition can occur over many hours causing the individual images to differ in exposure and lighting conditions. Therefore, to give the appearance of a single seamless image a blending operation is necessary. The quality of this blending depends on the magnitude of discontinuity along the boundaries between the images. Often image boundaries, or seams, are first computed to minimize this transition. Current techniques based on the multi-labeling Graph Cuts method are too slow and memory intensive for panoramas many gigapixels in size. In this paper we present a multithreaded out-of-core seam computing technique that is fast, has a small memory footprint, and gives near perfect scaling up to the number of physical cores of our test system. With this method the time required to compute image boundaries for gigapixel imagery improves from many hours (or even days) to just a few minutes on commodity hardware while still producing boundaries with energy that is on-par, if not better, than Graph Cuts.

As the visualization field matures, an increasing number of general toolkits are developed to cover a broad range of applications. However, no general tool can incorporate the latest capabilities for all possible applications, nor can the user interfaces and workflows be easily adjusted to accommodate all user communities. As a result, users will often chose either substandard solutions presented in familiar, customized tools or assemble a patchwork of individual applications glued through ad-hoc scripts and extensive, manual intervention. Instead, we need the ability to easily and rapidly assemble the best-in-task tools into custom interfaces and workflows to optimally serve any given application community. Unfortunately, creating such meta-applications at the API or SDK level is difficult, time consuming, and often infeasible due to the sheer variety of data models, design philosophies, limits in functionality, and the use of closed commercial systems. In this paper, we present the ManyVis framework which enables custom solutions to be built both rapidly and simply by allowing coordination and communication across existing unrelated applications. ManyVis allows users to combine software tools with complementary characteristics into one virtual application driven by a single, custom-designed interface.

Topological information has proven very valuable in the analysis of scientific data. An important challenge that remains is presenting this highly abstract information in a way that it is comprehensible even if one does not have an in-depth background in topology. Furthermore, it is often desirable to combine the structural insight gained by topological analysis with complementary information, such as geometric information. We present an overview over methods that use metaphors to make topological information more accessible to non-expert users, and we demonstrate their applicability to a range of scientific data sets. With the increasingly complex output of exascale simulations, the importance of having effective means of providing a comprehensible, abstract overview over data will grow. The techniques that we present will serve as an important foundation for this purpose.

The placement of tasks in a parallel application on specific nodes of a supercomputer can significantly impact performance. Traditionally, this task mapping has focused on reducing the distance between communicating tasks on the physical network. This minimizes the number of hops that point-to-point messages travel and thus reduces link sharing between messages and contention. However, for applications that use collectives over sub-communicators, this heuristic may not be optimal. Many collectives can benefit from an increase in bandwidth even at the cost of an increase in hop count, especially when sending large messages. For example, placing communicating tasks in a cube configuration rather than a plane or a line on a torus network increases the number of possible paths messages might take. This increases the available bandwidth which can lead to significant performance gains.

We have developed Rubik, a tool that provides a simple and intuitive interface to create a wide variety of mappings for structured communication patterns. Rubik supports a number of elementary operations such as splits, tilts, or shifts, that can be combined into a large number of unique patterns. Each operation can be applied to disjoint groups of processes involved in collectives to increase the effective bandwidth. We demonstrate the use of Rubik for improving performance of two parallel codes, pF3D and Qbox, which use collectives over sub-communicators.

Topological techniques have proven highly successful in analyzing and visualizing scientific data. As a result, significant efforts have been made to compute structures like the Morse-Smale complex as robustly and efficiently as possible. However, the resulting algorithms, while topologically consistent, often produce incorrect connectivity as well as poor geometry. These problems may compromise or even invalidate any subsequent analysis. Moreover, such techniques may fail to improve even when the resolution of the domain mesh is increased, thus producing potentially incorrect results even for highly resolved functions. To address these problems we introduce two new algorithms: (i) a randomized algorithm to compute the discrete gradient of a scalar field that converges under refinement; and (ii) a deterministic variant which directly computes accurate geometry and thus correct connectivity of the MS complex. The first algorithm converges in the sense that on average it produces the correct result and its standard deviation approaches zero with increasing mesh resolution. The second algorithm uses two ordered traversals of the function to integrate the probabilities of the first to extract correct (near optimal) geometry and connectivity. We present an extensive empirical study using both synthetic and real-world data and demonstrates the advantages of our algorithms in comparison with several popular approaches.

The performance of massively parallel applications is often heavily impacted by the cost of communication among compute nodes. However, determining how to best use the network is a formidable task, made challenging by the ever increasing size and complexity of modern supercomputers. This paper applies visualization techniques to aid parallel application developers in understanding the network activity by enabling a detailed exploration of the flow of packets through the hardware interconnect. In order to visualize this large and complex data, we employ two linked views of the hardware network. The first is a 2D view, that represents the network structure as one of several simplified planar projections. This view is designed to allow a user to easily identify trends and patterns in the network traffic. The second is a 3D view that augments the 2D view by preserving the physical network topology and providing a context that is familiar to the application developers. Using the massively parallel multi-physics code pF3D as a case study, we demonstrate that our tool provides valuable insight that we use to explain and optimize pF3D’s performance on an IBM Blue Gene/P system.

Analysis and visualization of complex vector fields remain major challenges when studying large scale simulation of physical phenomena. The primary reason is the gap between the concepts of smooth vector field theory and their computational realization. In practice, researchers must choose between either numerical techniques, with limited or no guarantees on how they preserve fundamental invariants, or discrete techniques which limit the precision at which the vector field can be represented. We propose a new representation of vector fields that combines the advantages of both approaches. In particular, we represent a subset of possible streamlines by storing their paths as they traverse the edges of a triangulation. Using only a finite set of streamlines creates a fully discrete version of a vector field that nevertheless approximates the smooth flow up to a user controlled error bound. The discrete nature of our representation enables us to directly compute and classify analogues of critical points, closed orbits, and other common topological structures. Further, by varying the number of divisions (quantizations) used per edge, we vary the resolution used to represent the field, allowing for controlled precision. This representation is compact in memory and supports standard vector field operations.

Representing uncertainty when creating visualizations is becoming more indispensable to understand and analyze scientific data. Uncertainty may come from different sources, such as, ensembles of experiments or unavoidable information loss when performing data reduction. One natural model to represent uncertainty is to assume that each position in space instead of a single value may take on a distribution of values. In this paper we present a new volume rendering method using per voxel Gaussian mixture models (GMMs) as the input data representation. GMMs are an elegant and compact way to drastically reduce the amount of data stored while still enabling realtime data access and rendering on the GPU. Our renderer offers efficient sampling of the data distribution, generating renderings of the data that flicker at each frame to indicate high variance. We can accumulate samples as well to generate still frames of the data, which preserve additional details in the data as compared to either traditional scalar indicators (such as a mean or a single nearest neighbor down sample) or to fitting the data with only a single Gaussian per voxel. We demonstrate the effectiveness of our method using ensembles of climate simulations and MRI scans as well as the down sampling of large scalar fields as examples.

Keywords: Uncertainty Visualization, Volume Rendering, Gaussian Mixture Model, Ensemble Visualization

The ViSUS software framework was designed with the primary philosophy that the visualization of massive data need not be tied to specialized hardware or infrastructure. In other words, a visualization environment for large data can be designed to be lightweight, highly scalable and run on a variety of plat- forms or hardware. Moreover, if designed generally such an infrastructure can have a wide variety of applications, all from the same code base. Figure 19.1 details example applications and the major components of the ViSUS infrastructure. The components can be grouped into three major categories. First, a lightweight and fast out-of-core data management framework using multi- resolution space filling curves. This allows the organization of information in an order that exploits the cache hierarchies of any modern data storage architectures. Second, a data flow framework that allows data to be processed during movement. Processing massive datasets in their entirety would be a long and expensive operation which hinders interactive exploration. By designing new algorithms to fit within this framework, data can be processed as it moves. Third, a portable visualization layer which was designed to scale from mobile devices to powerwall displays with same code base. In this chapter we will describe the ViSUS infrastructure, as well as give practical examples of its use in real world applications.

Exploring and analyzing the temporal evolution of features in large-scale time-varying datasets is a common problem in many areas of science and engineering. One natural representation of such data is tracking graphs, i.e., constrained graph layouts that use one spatial dimension to indicate time and show the “tracks” of each feature as it evolves, merges or disappears. However, for practical data sets creating the corresponding optimal graph layouts that minimize the number of intersections can take hours to compute with existing techniques. Furthermore, the resulting graphs are often unmanageably large and complex even with an ideal layout. Finally, due to the cost of the layout, changing the feature definition, e.g. by changing an iso-value, or analyzing properly adjusted sub-graphs is infeasible. To address these challenges, this paper presents a new framework that couples hierarchical feature definitions with progressive graph layout algorithms to provide an interactive exploration of dynamically constructed tracking graphs. Our system enables users to change feature definitions on-the-fly and filter features using arbitrary attributes while providing an interactive view of the resulting tracking graphs. Furthermore, the graph display is integrated into a linked view system that provides a traditional 3D view of the current set of features and allows a cross-linked selection to enable a fully flexible spatio-temporal exploration of data. We demonstrate the utility of our approach with several large-scale scientific simulations from combustion science.