Wesley Kendall

A Study of Parallel Particle Tracing for Steady-State and Time-Varying Flow Fields

Particle tracing for streamline and path line generation is a common method of visualizing vector fields in scientific data, but it is difficult to parallelize efficiently because of demanding and widely varying computational and communication loads. In this paper we scale parallel particle tracing for visualizing steady and unsteady flow fields well beyond previously published results. We configure the 4D domain decomposition into spatial and temporal blocks that combine in-core and out-of-core execution in a flexible way that favors faster run time or smaller memory. We also compare static and dynamic partitioning approaches. Strong and weak scaling curves are presented for tests conducted on an IBM Blue Gene/P machine at up to 32 K processes using a parallel flow visualization library that we are developing. Datasets are derived from computational fluid dynamics simulations of thermal hydraulics, liquid mixing, and combustion.

Scalable parallel building blocks for custom data analysis

We present a set of building blocks that provide scalable data movement capability to computational scientists and visualization researchers for writing their own parallel analysis. The set includes scalable tools for domain decomposition, process assignment, parallel I/O, global reduction, and local neighborhood communicationtasks that are common across many analysis applications. The global reduction is performed with a new algorithm, described in this paper, that efficiently merges blocks of analysis results into a smaller number of larger blocks. The merging is configurable in the number of blocks that are reduced in each round, the number of rounds, and the total number of resulting blocks. We highlight the use of our library in two analysis applications: parallel streamline generation and parallel Morse-Smale topological analysis. The first case uses an existing local neighborhood communication algorithm, whereas the latter uses the new merge algorithm.

An image compositing solution at scale

The only proven method for performing distributed-memory parallel rendering at large scales, tens of thousands of nodes, is a class of algorithms called sort last. The fundamental operation of sort-last parallel rendering is an image composite, which combines a collection of images generated independently on each node into a single blended image. Over the years numerous image compositing algorithms have been proposed as well as several enhancements and rendering modes to these core algorithms. However, the testing of these image compositing algorithms has been with an arbitrary set of enhancements, if any are applied at all. In this paper we take a leading production-quality image compositing framework, IceT, and use it as a testing frame work for the leading image compositing algorithms of today. As we scale IceT to ever increasing job sizes, we consider the image compositing systems holistically, incorporate numerous optimizations, and discover several improvements to the process never considered before. We conclude by demonstrating our solution on 64K cores of the Intrepid Blue Gene/P at Argonne National Laboratories.

Simplified parallel domain traversal

Many data-intensive scientific analysis techniques require global domain traversal, which over the years has been a bottleneck for efficient parallelization across distributed- memory architectures. Inspired by MapReduce and other simplified parallel programming approaches, we have designed DStep, a flexible system that greatly simplifies efficient parallelization of domain traversal techniques at scale. In order to deliver both simplicity to users as well as scalability on HPC platforms, we introduce a novel two-tiered communication architecture for managing and exploiting asynchronous communication loads. We also integrate our design with advanced parallel I/O techniques that operate directly on native simulation output. We demonstrate DStep by performing teleconnection analysis across ensemble runs of terascale atmospheric CO2 and climate data, and we show scalability results on up to 65,536 IBM BlueGene/P cores.

Toward a General I/O Layer for Parallel-Visualization Applications

For large-scale visualization applications, the visualization community urgently needs general solutions for efficient parallel I/O. These parallel visualization solutions should center around design patterns and the related data-partitioning strategies, not file formats. From this respect, its feasible to greatly alleviate I/O burdens without reinventing the wheel. For example, BIL (Block I/O Layer), which implements such a pattern, has greatly accelerated I/O performance for large-scale parallel particle tracing, a pervasive but challenging use case.

Scalable Multi-variate Analytics of Seismic and Satellite-based Observational Data

Over the past few years, large human populations around the world have been affected by an increase in significant seismic activities. For both conducting basic scientific research and for setting critical government policies, it is crucial to be able to explore and understand seismic and geographical information obtained through all scientific instruments. In this work, we present a visual analytics system that enables explorative visualization of seismic data together with satellite-based observational data, and introduce a suite of visual analytical tools. Seismic and satellite data are integrated temporally and spatially. Users can select temporal ;and spatial ranges to zoom in on specific seismic events, as well as to inspect changes both during and after the events. Tools for designing high dimensional transfer functions have been developed to enable efficient and intuitive comprehension of the multi-modal data. Spread-sheet style comparisons are used for data drill-down as well as presentation. Comparisons between distinct seismic events are also provided for characterizing event-wise differences. Our system has been designed for scalability in terms of data size, complexity (i.e. number of modalities), and varying form factors of display environments.

Terascale data organization for discovering multivariate climatic trends

Current visualization tools lack the ability to perform full-range spatial and temporal analysis on terascale scientific datasets. Two key reasons exist for this shortcoming: I/O and postprocessing on these datasets are being performed in suboptimal manners, and the subsequent data extraction and analysis routines have not been studied in depth at large scales. We resolved these issues through advanced I/O techniques and improvements to current query-driven visualization methods. We show the efficiency of our approach by analyzing over a terabyte of multivariate satellite data and addressing two key issues in climate science: time-lag analysis and drought assessment. Our methods allowed us to reduce the end-to-end execution times on these problems to one minute on a Cray XT4 machine.

Accelerating and benchmarking radix-k image compositing at large scale

Radix-k was introduced in 2009 as a configurable image compositing algorithm. The ability to tune it by selecting k-values allows it to benefit more from pixel reduction and compression optimizations than its predecessors. This paper describes such optimizations in Radix-k, analyzes their effects, and demonstrates improved performance and scalability. In addition to bounding and run-length encoding pixels, k-value selection and load balance are regulated at run-time. Performance is systematically analyzed for an array of process counts, image sizes, and HPC and graphics clusters. Analyses are performed using compositing of synthetic images and also in the context of a complete volume renderer and scientific data. We demonstrate increased performance over binary swap and show that 64 megapixels can be composited at rates of 0.08 seconds, or 12.5 frames per second, at 32 K processes.

Querying for Feature Extraction and Visualization in Climate Modeling

The ultimate goal of data visualization is to clearly portray features relevant to the problem being studied. This goal can be realized only if users can effectively communicate to the visualization software what features are of interest. To this end, we describe in this paper two query languages used by scientists to locate and visually emphasize relevant data in both space and time. These languages offer descriptive feedback and interactive refinement of query parameters, which are essential in any framework supporting queries of arbitrary complexity. We apply these languages to extract features of interest from climate model results and describe how they support rapid feature extraction from large datasets.

Parallel visualization on leadership computing resources

Changes are needed in the way that visualization is performed, if we expect the analysis of scientific data to be effective at the petascale and beyond. By using similar techniques as those used to parallelize simulations, such as parallel I/O, load balancing, and effective use of interprocess communication, the supercomputers that compute these datasets can also serve as analysis and visualization engines for them. Our team is assessing the feasibility of performing parallel scientific visualization on some of the most powerful computational resources of the U.S. Department of Energys National Laboratories in order to pave the way for analyzing the next generation of computational results. This paper highlights some of the conclusions of that research.