E. Wes Bethel
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
E. Wes Bethel has a total of 107 co-authors. Here are the top co-authors by number of shared publications:
- Gunther H. Weber
Lawrence Berkeley National Laboratory USA
5 shared publications - Kesheng Wu
Lawrence Berkeley National Laboratory, California, USA
4 shared publications - Bernd Hamann
Department of Computer Science, University of California, Davis, USA
2 shared publications - Brad Whitlock
Intelligent Light
2 shared publications - Burlen Loring
Lawrence Berkeley National Laboratory
2 shared publications C Johnson
University of Utah
2 shared publications
Latest Publications
- 5 May 2022
- Published by Springer Nature
- p. 439-460
Proximity Portability and in Transit, M-to-N Data Partitioning and Movement in SENSEI
- 5 May 2022
- Published by Springer Nature
- p. 439-460
Proximity Portability and in Transit, M-to-N Data Partitioning and Movement in SENSEI
- 5 May 2022
- Published by Springer Nature
- p. 281-306
The SENSEI Generic In Situ Interface: Tool and Processing Portability at Scale
- 5 May 2022
- Published by Springer Nature
- p. 281-306
The SENSEI Generic In Situ Interface: Tool and Processing Portability at Scale
- 15 June 2020
- Published by Springer Nature
- Vol. 12151, 127-145
Shared-Memory Parallel Probabilistic Graphical Modeling Optimization: Comparison of Threads, OpenMP, and Data-Parallel Primitives
This work examines performance characteristics of multiple shared-memory implementations of a probabilistic graphical modeling (PGM) optimization code, which forms the basis for an advanced, state-of-the art image segmentation method. The work is motivated by the need to accelerate scientific image analysis pipelines in use by experimental science, such as at x-ray light sources, and is motivated by the need for platform-portable codes that perform well across many different computational architectures. The primary focus of this work and its main contribution is an in-depth study of shared-memory parallel performance of different implementations, which include those using alternative parallelization approaches such as C11-threads, OpenMP, and data parallel primitives (DPPs). Our results show that, for this complex data-intensive algorithm, the DPP implementation exhibits better runtime performance, but also exhibits less favorable scaling characteristics than the C11-threads and OpenMP counterparts. Based upon a set of experiments that collect hardware performance counters on multiple platforms, the reason for the runtime performance difference appears to be due primarily to algorithmic efficiency gains: the reformulation from the traditional C11-threads and OpenMP expression of the solution into that of data parallel primitives results in significantly fewer instructions being executed. This study is the first of its type to do performance analysis using hardware counters for comparing methods based on VTK-m-based data-parallel primitives with those based on more traditional OpenMP or threads-based parallelism. It is timely, as there is increasing awareness of the need for platform portability in light of increasing node-level parallelism and increasing device heterogeneity.
- 15 June 2020
- Published by Springer Nature
- Vol. 12151, 127-145
Shared-Memory Parallel Probabilistic Graphical Modeling Optimization: Comparison of Threads, OpenMP, and Data-Parallel Primitives
This work examines performance characteristics of multiple shared-memory implementations of a probabilistic graphical modeling (PGM) optimization code, which forms the basis for an advanced, state-of-the art image segmentation method. The work is motivated by the need to accelerate scientific image analysis pipelines in use by experimental science, such as at x-ray light sources, and is motivated by the need for platform-portable codes that perform well across many different computational architectures. The primary focus of this work and its main contribution is an in-depth study of shared-memory parallel performance of different implementations, which include those using alternative parallelization approaches such as C11-threads, OpenMP, and data parallel primitives (DPPs). Our results show that, for this complex data-intensive algorithm, the DPP implementation exhibits better runtime performance, but also exhibits less favorable scaling characteristics than the C11-threads and OpenMP counterparts. Based upon a set of experiments that collect hardware performance counters on multiple platforms, the reason for the runtime performance difference appears to be due primarily to algorithmic efficiency gains: the reformulation from the traditional C11-threads and OpenMP expression of the solution into that of data parallel primitives results in significantly fewer instructions being executed. This study is the first of its type to do performance analysis using hardware counters for comparing methods based on VTK-m-based data-parallel primitives with those based on more traditional OpenMP or threads-based parallelism. It is timely, as there is increasing awareness of the need for platform portability in light of increasing node-level parallelism and increasing device heterogeneity.
- 3 April 2012
- Published by SAGE Publications in The International Journal of High Performance Computing Applications
- Vol. 26 (4) , 399-412
Multi-core and many-core shared-memory parallel raycasting volume rendering optimization and tuning
Given the computing industry trend of increasing processing capacity by adding more cores to a chip, the focus of this work is tuning the performance of a staple visualization algorithm, raycasting volume rendering, for shared-memory parallelism on multi-core CPUs and many-core GPUs. Our approach is to vary tunable algorithmic settings, along with known algorithmic optimizations and two different memory layouts, and measure performance in terms of absolute runtime and L2 memory cache misses. Our results indicate there is a wide variation in runtime performance on all platforms, as much as 254% for the tunable parameters we test on multi-core CPUs and 265% on many-core GPUs, and the optimal configurations vary across platforms, often in a non-obvious way. For example, our results indicate the optimal configurations on the GPU occur at a crossover point between those that maintain good cache utilization and those that saturate computational throughput. This result is likely to be extremely difficult to predict with an empirical performance model for this particular algorithm because it has an unstructured memory access pattern that varies locally for individual rays and globally for the selected viewpoint. Our results also show that optimal parameters on modern architectures are markedly different from those in previous studies run on older architectures. In addition, given the dramatic performance variation across platforms for both optimal algorithm settings and performance results, there is a clear benefit for production visualization and analysis codes to adopt a strategy for performance optimization through auto-tuning. These benefits will likely become more pronounced in the future as the number of cores per chip and the cost of moving data through the memory hierarchy both increase.
- 3 April 2012
- Published by SAGE Publications in The International Journal of High Performance Computing Applications
- Vol. 26 (4) , 399-412
Multi-core and many-core shared-memory parallel raycasting volume rendering optimization and tuning
Given the computing industry trend of increasing processing capacity by adding more cores to a chip, the focus of this work is tuning the performance of a staple visualization algorithm, raycasting volume rendering, for shared-memory parallelism on multi-core CPUs and many-core GPUs. Our approach is to vary tunable algorithmic settings, along with known algorithmic optimizations and two different memory layouts, and measure performance in terms of absolute runtime and L2 memory cache misses. Our results indicate there is a wide variation in runtime performance on all platforms, as much as 254% for the tunable parameters we test on multi-core CPUs and 265% on many-core GPUs, and the optimal configurations vary across platforms, often in a non-obvious way. For example, our results indicate the optimal configurations on the GPU occur at a crossover point between those that maintain good cache utilization and those that saturate computational throughput. This result is likely to be extremely difficult to predict with an empirical performance model for this particular algorithm because it has an unstructured memory access pattern that varies locally for individual rays and globally for the selected viewpoint. Our results also show that optimal parameters on modern architectures are markedly different from those in previous studies run on older architectures. In addition, given the dramatic performance variation across platforms for both optimal algorithm settings and performance results, there is a clear benefit for production visualization and analysis codes to adopt a strategy for performance optimization through auto-tuning. These benefits will likely become more pronounced in the future as the number of cores per chip and the cost of moving data through the memory hierarchy both increase.
- 31 March 2012
- Published by With Intelligence LLC in The Journal of Trading
- Vol. 7 (2) , 9-25
Federal Market Information Technology in the Post–FlashCrash Era: Roles for Supercomputing
This article describes collaborative work between active traders, regulators, economists, and supercomputing researchers to replicate and extend investigations of the Flash Crash and other market anomalies in a National Laboratory high-performance computing (HPC) environment. Our work suggests that supercomputing tools and methods will be valuable to market regulators in achieving the goals of market safety, stability, and security. Research results using high-frequency data and analytics are described, and directions for future development are discussed. Currently, the key mechanism for preventing catastrophic market action is “circuit breakers.” We believe a more graduated approach, similar to the “yellow light” approach in motorsports to slow down traffic, might be a better way to achieve the same goal. To enable this objective, we study a number of indicators that could foresee hazards in market conditions and explore options to confirm such predictions. Our tests confirm that volume-synchronized probability of informed trading (VPIN) and a version of the volume Herfindahl–Hirschman index (HHI) for measuring market fragmentation could have indeed given strong signals ahead of the Flash Crash event on May 6, 2010. This is a preliminary step toward a full-fledged earlywarning system for unusual market conditions.
- 31 March 2012
- Published by With Intelligence LLC in The Journal of Trading
- Vol. 7 (2) , 9-25
Federal Market Information Technology in the Post–FlashCrash Era: Roles for Supercomputing
This article describes collaborative work between active traders, regulators, economists, and supercomputing researchers to replicate and extend investigations of the Flash Crash and other market anomalies in a National Laboratory high-performance computing (HPC) environment. Our work suggests that supercomputing tools and methods will be valuable to market regulators in achieving the goals of market safety, stability, and security. Research results using high-frequency data and analytics are described, and directions for future development are discussed. Currently, the key mechanism for preventing catastrophic market action is “circuit breakers.” We believe a more graduated approach, similar to the “yellow light” approach in motorsports to slow down traffic, might be a better way to achieve the same goal. To enable this objective, we study a number of indicators that could foresee hazards in market conditions and explore options to confirm such predictions. Our tests confirm that volume-synchronized probability of informed trading (VPIN) and a version of the volume Herfindahl–Hirschman index (HHI) for measuring market fragmentation could have indeed given strong signals ahead of the Flash Crash event on May 6, 2010. This is a preliminary step toward a full-fledged earlywarning system for unusual market conditions.
- 1 January 2012
- Published by Springer Nature
- p. 265-283
Linking Advanced Visualization and MATLAB for the Analysis of 3D Gene Expression Data
- 1 January 2012
- Published by Springer Nature
- p. 265-283
Linking Advanced Visualization and MATLAB for the Analysis of 3D Gene Expression Data
- 27 May 2010
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Visualization and Computer Graphics
- Vol. 17 (3) , 264-275
An Application of Multivariate Statistical Analysis for Query-Driven Visualization
Driven by the ability to generate ever-larger, increasingly complex data, there is an urgent need in the scientific community for scalable analysis methods that can rapidly identify salient trends in scientific data. Query-Driven Visualization (QDV) strategies are among the small subset of techniques that can address both large and highly complex data sets. This paper extends the utility of QDV strategies with a statistics-based framework that integrates nonparametric distribution estimation techniques with a new segmentation strategy to visually identify statistically significant trends and features within the solution space of a query. In this framework, query distribution estimates help users to interactively explore their query's solution and visually identify the regions where the combined behavior of constrained variables is most important, statistically, to their inquiry. Our new segmentation strategy extends the distribution estimation analysis by visually conveying the individual importance of each variable to these regions of high statistical significance. We demonstrate the analysis benefits these two strategies provide and show how they maybe used to facilitate the refinement of constraints over variables expressed in a user's query. We apply our method to data sets from two different scientific domains to demonstrate its broad applicability.
- 27 May 2010
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Visualization and Computer Graphics
- Vol. 17 (3) , 264-275
An Application of Multivariate Statistical Analysis for Query-Driven Visualization
Driven by the ability to generate ever-larger, increasingly complex data, there is an urgent need in the scientific community for scalable analysis methods that can rapidly identify salient trends in scientific data. Query-Driven Visualization (QDV) strategies are among the small subset of techniques that can address both large and highly complex data sets. This paper extends the utility of QDV strategies with a statistics-based framework that integrates nonparametric distribution estimation techniques with a new segmentation strategy to visually identify statistically significant trends and features within the solution space of a query. In this framework, query distribution estimates help users to interactively explore their query's solution and visually identify the regions where the combined behavior of constrained variables is most important, statistically, to their inquiry. Our new segmentation strategy extends the distribution estimation analysis by visually conveying the individual importance of each variable to these regions of high statistical significance. We demonstrate the analysis benefits these two strategies provide and show how they maybe used to facilitate the refinement of constraints over variables expressed in a user's query. We apply our method to data sets from two different scientific domains to demonstrate its broad applicability.
- 1 July 2009
- Published by IOP Publishing in Journal of Physics: Conference Series
- Vol. 180 (1) , 012084
Occam's razor and petascale visual data analysis
One of the central challenges facing visualization research is how to effectively enable knowledge discovery. An effective approach will likely combine application architectures that are capable of running on today's largest platforms to address the challenges posed by large data with visual data analysis techniques that help find, represent, and effectively convey scientifically interesting features and phenomena.
- 1 July 2009
- Published by IOP Publishing in Journal of Physics: Conference Series
- Vol. 180 (1) , 012084
Occam's razor and petascale visual data analysis
One of the central challenges facing visualization research is how to effectively enable knowledge discovery. An effective approach will likely combine application architectures that are capable of running on today's largest platforms to address the challenges posed by large data with visual data analysis techniques that help find, represent, and effectively convey scientifically interesting features and phenomena.