Communication performance is of paramount importance to high performance computing (HPC) applications. MPI is widely used in HPC due to its portability across platforms. However, the performance of MPI implementations on large-scale supercomputers is significantly impacted by factors including its inherent buffering, type checking, and other control overheads. 

At the same time, we are witnessing the advent of intelligent network interfaces and system interconnects together with support for paradigms such as active messages. The Parallel Active Message Interface (PAMI) library for the IBM Blue Gene/Q system and the User Generic Network Interface (uGNI) for the Cray XE6 system are communication libraries enabling one to fully exploit the features of the underlying network interconnects. In this poster, we evaluate the performance of MPI with PAMI and uGNI and demonstrate that these libraries achieve significant improvements in communication performance over MPI. 

We are working towards creating abstractions that hide the complexities found in current system interconnects, and expose a simple API to enable applications and middleware developers fully exploit the underlying features.

SC12, Salt Lake City, Utah

Evaluating Communication Performance of BlueGene/Q and Cray XE6 Supercomputers

As the scale and complexity of data continue to grow at unprecedented rates, scientists are increasingly relying on Large, High-Resolution Displays to visualize and analyze scientific datasets. Recent studies have demonstrated the effectiveness of these displays in supporting cognitively demanding data analysis and sensemaking tasks. While there has been an abundance of research on rendering algorithms for large, high-resolution displays, far less effort has gone into designing interactive visual analytic interfaces to effectively leverage these displays in visual exploration and sensemaking scenarios involving large collections of data. In this paper, we present an interactive visual analytics application for the exploration of large trajectory datasets. Our application utilizes large, high-resolution 3D display environments to simultaneously visualize and juxtapose a large number of trajectories. It also integrates a scalable visual query technique, which can be used to quickly formulate and verify hypotheses, encouraging scientists to contemplate multiple competing theories before drawing conclusions. We evaluate our design within the context of a behavioral ecology case study. We also share our observations from a pilot user study to provide insights on how scientists might utilize large display environments in visual exploration and sensemaking scenarios. 

Keywords: Large, High-Resolution displays; visual exploration; sensemaking; trajectory analysis

Trajectory analysis application running on a 18 Megapixel, thin-bezel, tiled 3D display wall at the Electronic Visualization Laboratory

Scalable Visual Queries for Data Exploration on Large, High-Resolution 3D Displays

Current techniques in medical imaging and analysis primarily focus on recording information about one specific physiological property at a time. Various modalities such as magnetic resonance, computed tomography, and digital subtraction angiography are each suited towards different tasks. In order to improve surgical planning, physicians would benefit from patient specific computational models built from medical images. These models could be used in order to run simulations and simultaneously gather physiological information that would otherwise require multiple imaging modalities or be impossible to measure with current technology. 

We present a pipeline for processing medical data and executing computational simulations to enhance the information conveyed in standard medical imaging. Our work focuses on the whole brain, where we&rsquo;ve developed tools that allow vasculature to be analyzed in three-dimensions, at high resolutions, and with multiple relevant data sets overlaid on the vascular structure. In order to avoid confusion and misinterpretations, we have the ability to render simulated data such that it mirrors raw medical images and vascular reconstructions. 

Keywords: Medical image reconstruction, computer simulation, voxelization, vasculature 

Index Terms: I.4.1 [Image Processing and Computer Vision]: Digitization and Image Capture - Imaging Geometry; J.3 [Life and Medical Science]: Health

Scalar values visualized in color to allow for quick interpretation of  the vascular network data

Brain Vascular Reconstruction, Simulation, and Visualization

Objective: To test the hypothesis that Hands-on Automated Nursing Data System (HANDS) &ldquo;big picture summary&rdquo; can be implemented uniformly across diverse settings, and result in positive registered nurse (RN) and plan of care (POC) data outcomes across time. 
Design: In a longitudinal, multisite, full test study, a representative convenience sample of eight medical-surgical units from four hospitals (one university, two large community, and one small community) in one Midwestern state implemented the HANDS intervention for 24 (four units) or 12 (four units) months. 
Measurements: (a) RN outcomes-percentage completing training, satisfaction with standardized terminologies, perception of HANDS usefulness, POC submission compliance rate. (b) POC data outcomes-validity (rate of optional changes/episode); reliability of terms and ratings; and volume of standardized data generated. 
Results: One hundred percent of the RNs who worked on the eight study units successfully completed the required standardized training; all units selected participated for the entire 12- or 24-month designated period; compliance rates for POC entry at every patient hand-off were 78-92%; reliability coefficients for use of the standardized terms and ratings were moderately strong; the pattern of optional POC changes per episode declined but remained reasonable across time; and the nurses generated a database of 40,747 episodes of care. 
Limitations: Only RNs and medical-surgical units participated. 
Conclusion: It is possible to effectively standardize the capture and visualization of useful &ldquo;big picture&rdquo; healthcare information across diverse settings. Findings offer a viable alternative to the current practice of introducing new health information layers that ultimately increase the complexity and inconsistency of information for frontline users.

The HANDS Ecosystem.

Maintaining a consistent big picture: Meaningful Use of a Web-based POC EHR System

In visualization of tensor data associated with turbulent combustion simulations, the datasets we have to process and visualize are typically enormous in size and continue to grow exponentially. The large amount of data may severely affect the manipulating speed, thus posing a major challenge to interactive visualization. Even worse, tensor datasets tend to be very dense leading to clutter and occlusion problems. To address these issues, feature extraction is an emerging method. Typically, only a small percentage of data is of interest, thus making the effective visualization of very large datasets possible. Feature extraction also helps the users highlight and focus on regions of interest. 
We introduce an approach for the segmentation, visualization and tracking of regions of interest in large scale tensor field datasets generated by computational turbulent combustion simulations. We use canopy clustering followed by a K-means algorithm to partition and cluster the tensor field components. The resulting clusters are tracked through multiple timesteps. Interactive, hardware-accelerated volume renderings are generated using the cluster indices. Results on two rich datasets show this approach can assist in the visual analysis of combustion tensor fields. 
A great deal of research has been conducted on the problems of feature extraction and tracking, as well as in characterizing, detecting and visualizing regions of interes. The method we present herein is novel in that it integrates machine learning with visualization for extracting and clustering regions of interest. It is thus a promising approach to apply to very large flow datasets.

Tensor-based region of interest tracking for a mixing layer dataset.

A Novel Method for Tracking Tensor-based Regions of Interest in Large-Scale, Spatially-Dense Turbulent Combustion Data

This paper examines the role of media arts in the development of interactive learning environments for science, technology, engineering, and mathematics (STEM). The use of technology-mediated interactive environments already provides greater access for science learning to wider, more diverse, and especially younger audiences. Since children typically enjoy playing computer games and interacting with new technology, media art provides a motivating setting for science education. The movement to integrate science and art to examine convergence points, and how the fields can empower each other, has already begun. However, the role of arts and artists in the design of advanced applications for STEM education has yet to become subject for research and implementation in mainstream cultural institutions and educational programs. Collaborative, interdisciplinary teams of artists, technologists and scientists developed novel interactive learning projects that educate the public on fundamental science (STEM) disciplines. Five interactive educational applications were designed based on the leading art and design concepts with a focus on user engagement, interactive design, and aesthetics principles. With these principles at their center, these applications and their educational content were designed to create interactive and engaging learning experiences. We describe the role of art in the development of these projects and examine how artists can cross disciplinary borders to collaborate in the development of innovative educational STEM learning applications. 

Keywords: Media Arts, STEM, Science Education, art-science, visualization, interactive applications

Users experiencing A Mile in My Paws immersive environment.

Media Arts in Support of Science Education, ISEA2012

This OpenGL graphics engine is my first graphics engine. It can be used to view complex geometric models as well as simple primitives. The goal of this project was to familiarize myself with OpenGL since in the past I have always used tools that hide the underlying OpenGL code. The graphics engine has a number of features such as Wavefront object file support, Phong shading model, basic shadows, interlaced 3D support, 2D UI support, built in camera controls, as well as a simple API to create custom scenes. This paper will describe the functionality of the graphics engine as well as the tools used and development process.

A scene with multiple light sources being rendered with the Phong vertex and fragment shaders

OpenGL Graphics Engine

We evaluate the use of Poisson reconstruction to generate a 3D bathymetric model of West Lake Bonney, Antarctica. The source sonar dataset has been collected by the ENDURANCE autonomous vehicle in the course of two Antarctic summer missions. The reconstruction workflow involved processing 200 million datapoints to generate a high resolution model of the lake bottom, Narrows region and underwater glacier face. A novel and flexible toolset has been developed to automate the processing of the Bonney data.

Side-by-side analysis of West Lake Bonney reconstruction results on a high resolution display wall

Poisson reconstruction of extreme submersed environments: The ENDURANCE exploration of an under-ice Antarctic Lake

This demo presents AstroShelf, our on-going effort to enable astrophysicists to collaboratively investigate celestial objects using data originating from multiple sky surveys, hosted at different sites. The AstroShelf platform combines database and data stream, workflow and visualization technologies to provide a means for querying and displaying telescope images (in a Google Sky manner), visualizations of spectrum data, and for managing annotations. In addition to the user interface, AstroShelf supports a programmatic interface (available as a web service), which allows astrophysicists to incorporate functionality from AstroShelf in their own programs. A key feature is Live Annotations which is the detection and delivery of events or annotations to users in real-time, based on their profiles. We demonstrate the capabilities of AstroShelf through real end user exploration scenarios (with participation from &ldquo;stargazers&rdquo; in the audience), in the presence of simulated annotation workloads executed through web services. 
Categories and Subject Descriptors: H.2.8 [DATABASE MANAGEMENT]: Database Applications-
Scientific databases; H.3.5 [INFORMATION STORAGE AND RETRIEVAL]: Online Information Services-Data Sharing, Web-Based Services 
Keywords: Big Data, Scientific Databases, Continuous Workflows, Annotations, Visualization, Astronomy

Screenshot from AstroShelf’s User Interface, currently showing the SkyView (left) and the Live Annotations View (right).

AstroShelf: Understanding the Universe Through Scalable Navigation of a Galaxy of Annotations

This work presents a novel immersive workstation environment that scientists can use for 3D data exploration and as their everyday 2D computer monitor. This implementation is based on an autostereoscopic dynamic parallax barrier 2D / 3D display, interactive input devices. This paper will describe the my contributions to the hardware construction, software components, and its implementation as a platform for nanoscale materials science exploration. 

Keywords: Information interfaces and presentation; Multimedia information systems; Artificial, augmented, and virtual realities

User multitouch interaction with Dynallax

Transforming Dynallax Into A Simultaneous 2D / 3D AutostereoWorkstation

We report findings on the current state of pain care in hospitals for end-of-life (EOL) patients using longitudinal data from eight diverse medical-surgical units located in 4 different Midwestern hospitals over 24 months. We identified 1,425 EOL care episodes, 596 (41.3%) of which had a pain diagnosis. The percentage of EOL patients with pain varied significantly across units (p<.001), and was even lower (27.7%) for those with &ldquo;acute confusion.&rdquo; Additionally, 30% of EOL patients had severe or significant pain at death or discharge to hospice and only 42.7% actually met the expected pain related outcome ratings. Pain often improved within 48 hours of admission (p<.005), the improvement, however, stagnated following this initial time period (p=.92). A sizable gap between pain science and clinical practice continues.

Actual final Nursing Outcomes Classification (NOC) outcome ratings for end-of-life patients with pain diagnoses.

Current State of Pain Care for Hospitalized Patients at End of Life

This article will describe the progress since 2000 on research and development in 2-D and 3-D scalable resolution display walls that are built from tiling individual lower resolution flat panel displays. The article will describe approaches and trends in display hardware construction, middleware architecture, and user-interaction design. The article will also highlight examples of use cases and the benefits the technology has brought to their respective disciplines. 

Keywords: Computer graphics; stereo vision; virtual reality; visualization

A user partitions on the work space in SAGE by drawing wall dividers with his finger

Scalable Resolution Display Walls

IEEE LDAV'12 Best-Paper Runner-Up (Honorable Mention) 
We introduce a web-based, client-server computing infrastructure to assist the interactive navigation of large-scale astronomy observations. Large image datasets are partitioned into a spatial index structure that allows prefix-matching of spatial objects. In conjunction with pixel-based overlays, this approach allows fetching, displaying, panning and zooming of gigabit panoramas of the sky in real time. Images from three sky surveys (SDSS, FIRST and simulated LSST results) are cross-registered and integrated as overlays, allowing cross-spectrum analysis of astronomy observations. The front-end of the infrastructure uses the web technologies WebGL and HTML5 to enable cross-platform, web-based functionality. Our approach attains interactive rendering framerates; its power and flexibility enables us to serve the needs of the astronomy community. Evaluation on a galaxy case study, as well as feedback from domain experts emphasize the benefits of this visual approach to the observational astronomy field. 
Index Terms: Data Fusion and Integration, Scalability Issues, Zooming and Navigation Techniques, Geographic/Geospatial Visualization

Four decreasing level-of-detail versions of a sky image tile (for demonstration purposes, an SDSS image).

Panning and Zooming the Observable Universe with Preﬁx-Matching Indices and Pixel-Based Overlays

While many visualization tools exist that offer sophisticated functions for charting complex data, synthesizing various information and deriving insight from abstract data, they still expect users to possess a high degree of expertise in wielding the tools to create an effective visualization. Unfortunately the users of such tools are usually lay people or domain experts with marginal knowledge of visualization techniques. These users typically can&rsquo;t make use of the latest advances in visualization tools because of the high learning curve, instead they want to focus on the data, and often end up resorting to traditional bar chart or spreadsheet. To facilitate the use of the advanced visualization tools, I propose an automated visualization framework: Articulate. The goal is to provide a streamlined experience to non-expert users, allowing them to focus on using the visualizations effectively to generate new findings. 

The 2008 National Science Foundation workshop report on &ldquo;Enabling Science Discoveries through Visual Exploration&rdquo; also noted &ldquo;there is a strong desire for conversational interfaces that facilitate a more natural means of interacting with science.&rdquo; Even a decade ago this would have seemed far-fetched, but today there is renewed interest in the use of natural language as an interface to computing, with a variety of successful models that are able to understand the meaning of verbal descriptions in search engine, recommender system, educational technology and health applications. This inspired me to adopt a natural language interface in the automated visualization framework which would allow an end-user to pose verbal inquiries, and then let the system assume the burden of determining the appropriate graph, and presenting the results. It is hoped that such a capability can potentially reduce the learning curve necessary for effective use of visualization tools, and thereby expand the population of users who can successfully conduct visual analysis. 

In this dissertation I present Articulate, an approach toward enabling non-visualization experts to leverage advanced visualization techniques through the use of natural language as the primary interface for crafting visualizations. The main challenge in this research is in determining how to translate imprecise verbal queries into precise and meaningful visualizations. In brief, the process involves three main steps: first, extracting syntactic and semantic information from a verbal query; then applying a supervised learning algorithm to automatically translate the user&rsquo;s intention into explicit expressions; and finally, selecting a suitable visualization to depict the expression. While the initial prototype produces information visualizations, it will also be demonstrated how the approach can be applied to scientific visualization as well as the details of the user studies that validate the approach. 

It is important to understand that the goal of this research is to explore how to automatically translate natural and potentially ill-defined language into meaningful visualizations of data in a generalizable way, in order to relieve the user of the burden of having to learn how to use a complex interface, so that they can focus on articulating better scientific questions in data exploration with the modern advances of visualization.

Screen Shot of Articulate&rsquo;s User Interface

Articulate: Creating Meaningful Visualizations from Natural Language

In the CineGrid Exchange project, we use iRODS to curate and preserve super-high-definition media content such as 4K movies. Our media curation platform enables long-term data preservation and streamlined workflows for curators to handle large amount of media data in a cross-organizational environment. The platform consists of a front-end web catalog system and a back-end storage and workflow system. 

The front-end is a web-based content management system and brings ease-of-use in metadata management and data retrieval. The back-end is an iRODS-based distributed storage system that guarantees data consistency and redundancy. It also provides an event-based rule engine to realize flexible workflows.

Back-end storage and workflow system using iRODS

Global Media Curation Platform Using iRODS

Implementation: 
<ul>
<li>Next generation of the Dynallax: a dynamic parallax barrier autostereoscopy system in a desktop form factor.</li>
<li>It is driven by a software system called Dynamic View Client (DVC) which consist of two parts: 
<ul>
<li>DVCSdk: A server-side library that enables streaming of stereo image pairs to the client side and interacts with applications to process client input</li>
<li>DVCLib: A client-side library that displays stereo images, supports interaction devices, and provides a simultaneous 2D and 3D workspace.</li>
</ul>
</li>
</ul>

Objectives: 
<ul>
<li>Deploy a transparent virtual environment that directs the user&rsquo;s focus on the science problem instead of the technology.</li>
<li>Support existing workflows so that scientists need not give up tools upon which they rely to be productive.</li>
<li>Design an accessible environment that is always available and requires little or no added effort on the part of users and developers.</li>
</ul>

User Multitouch Interaction with Dynallax

A Simultaneous 2D / 3D Autostereo Workstation (poster)

We present a novel immersive workstation environment that scientists can use for 3D data exploration and as their everyday 2D computer monitor. Our implementation is based on an autostereoscopic dynamic parallax barrier 2D / 3D display, interactive input devices, and a software infrastructure that allows client / server software modules to couple the workstation to scientists&rsquo; visualization applications. This paper describes the hardware construction and calibration, software components, and a demonstration of our system in nanoscale materials science exploration. 

Keywords: Information interfaces and presentation; Multimedia information systems; Artificial, augmented, and virtual realities.

The Dynallax display system

A Simultaneous 2D / 3D Autostereo Workstation

The primary goal of visualization is insight. An effective visualization is best achieved through the creation of a proper representation of data and the interactive manipulation and querying of the visualization. Large-scale data visualization is particularly challenging because the size of the data is several orders of magnitude larger than what can be managed on an average desktop computer. Data sizes range from terabytes to petabytes (and soon exabytes) rather than a few megabytes to gigabytes. Large-scale data can also be of much greater dimensionality, and there is often a need to correlate it with other types of similarly large and complex data. Furthermore the need to query data at the level of individual data samples is superseded by the need to search for larger trends in the data. Lastly, while interactive manipulation of a derived visualization is important, it is much more difficult to achieve because each new visualization requires either re-traversing the entire dataset, or compromising by only viewing a small subset of the whole. Large-scale data visualization therefore requires the use of distributed computing. The individual components of a data visualization pipeline can be abstracted as: 
Data Retrieval -> Filter / Mine -> Render -> Display 
The degree to which these individual components are distributed or collocated has historically been driven by the cost to deploy and maintain infrastructure and services. Early in the history of scientific computing, networking bandwidth was expensive and therefore scarce. Consequently early visualization pipelines tended to minimize the movement of data over networks in favor of collocating data storage with data processing. However, as the amount and variety of data continued to grow at an exponential pace, it became too costly to maintain full replicas of the data for each individual that needed to use it. Instead, by leveraging the widespread expansion of the Internet and other national and international high-speed network infrastructure such as the National LambdaRail, Internet-2, and the Global Lambda Integrated Facility, data and service providers began to migrate toward a model of widespread distribution of resources. 
In this chapter we will first introduce the various instantiations of the visualization pipeline and the historic motivation for their creation. We will then examine individual components of the pipeline in detail to understand the technical challenges that must be solved in order to ensure continued scalability. We will discuss distributed data management issues that are specifically relevant to large-scale visualization. We will also introduce key data rendering techniques and explain through case studies approaches for scaling them by leveraging distributed computing. Lastly we will describe advanced display technologies that are now considered the &ldquo;lenses&rdquo; for examining large-scale data.

Visualization of Large-Scale Distributed Data

Simulation and modeling of turbulent flow, and of turbulent reacting flow in particular, involves solving for and analyzing time-dependent and spatially dense tensor quantities, such as turbulent stress tensors. The interactive visual exploration of these tensor quantities can effectively steer the computational modeling of combustion systems. In this chapter, we discuss the challenges in dense symmetric tensor visualization applied to turbulent combustion calculation, and analyze the feasibility of using several established tensor visualization techniques in the context of exploring space-time relationships in computationally-simulated combustion tensor data. To tackle the pervasive problems of occlusion and clutter, we propose a solution combining techniques from information and scientific visualization. Specifically, the proposed solution combines a detailed 3D inspection view based on volume rendering with glyph-based representations - used as 2D probes -, while leveraging interactive filtering and flow salience cues to clarify the structure of the tensor datasets. Side-by-side views of multiple timesteps facilitate the analysis of time-space relationships. The resulting prototype enables an analysis style based on the overview first, zoom and filter, then details on demand paradigm originally proposed in information visualization. The result is a visual analysis tool to be utilized in debugging, benchmarking, and verification of models and solutions in turbulent combustion. We demonstrate this analysis tool on three example configurations and report feedback from combustion researchers.

DNS (left), LES (middle) and RANS (right) predictions of a turbulent jet. LES requires less computational effort than DNS, while delivering more detail than the inexpensive RANS.