We propose and discuss a paradigm that allows for expressing data-parallel rendering with the classically non-parallel ANARI API. We propose this as a new standard for data-parallel rendering, describe two different i...
详细信息
ISBN:
(纸本)9798331516932;9798331516925
We propose and discuss a paradigm that allows for expressing data-parallel rendering with the classically non-parallel ANARI API. We propose this as a new standard for data-parallel rendering, describe two different implementations of this paradigm, and use multiple sample integrations into existing applications to show how easy it is to adopt, and what can be gained from doing so.
To analyze large amounts of numerical data, one of the most useful approaches is to use scientific visualization to transform them into graphical images. Flow visualization as one of the challenging topics has played ...
详细信息
ISBN:
(纸本)9781424433940
To analyze large amounts of numerical data, one of the most useful approaches is to use scientific visualization to transform them into graphical images. Flow visualization as one of the challenging topics has played important roles in oceanic dataanalysis. There are many techniques have been presented in the past decade, but most of them can't get high performance to visualize large-scale flow data in real time. To deduce the computational complexity brought by large flow dataset, feature-based expression will be a helpful way. However, how to get the result images quickly without costing much time for feature extraction and analysis is a very important problem to deal. Based on the common characteristic of flow and the unchangeable scale feather of spiral line, we present a new distributing strategy which needn't locate feature points very accurately and didn't rely on the type of feature fields. The visualization procedure not only can straight forward automatically but also can be changed with user's interactive command. The flow data obtained from the South Sea of China was verified and simulated. The result shows that this method using spiral strategy not templates to setting the seeds to emphasize the interesting fields is much faster and flexible, especially in large-scale flow datavisualization
One of the greatest challenges facing scientists doing large computation of vector fields in a distributed parallel setting is the need for optimal parallel algorithms for flow visualization. To address this need, we ...
详细信息
ISBN:
(纸本)9781538668733
One of the greatest challenges facing scientists doing large computation of vector fields in a distributed parallel setting is the need for optimal parallel algorithms for flow visualization. To address this need, we present a new flow visualization method based on parallel 3D line integral convolution (LIC). Our approach uses the fact that 3D LIC only needs limited local information to design an embarrassingly parallel model with a trade-off between the additional memory cost of external cells and the time cost for communication. All data required for each process can be stored in either a local data block or the external cells which are sets of exterior data surrounding the local partition. One problem for parallel LIC is that equal domain size decomposition of the data cannot guarantee balanced parallel processes. To achieve a load-balanced visualization process, we repartition data using an estimate of the LIC computation time. In addition, to minimize the memory cost, we introduce a vector-driven external cell expansion method to reduce the required memory cost. We find that we can use fewer external cells with minimal loss of visual quality. We evaluate the performance of our visualization method by first comparing its parallel scalability with traditional integral field line visualization. Next, we compare our new partition method with other data partition methods to verify that the workload of our model is more balanced. Finally, we compare our external cell expansion method with a traditional layer-based external cell expansion method. Consequently, together with the new partition and external cell expansion methods, our parallel 3D LIC visualization proves to be an efficient and well-balanced parallel flow visualization with limited extra memory cost and a large saving of communication time.
The primary challenge in the visualization of a large-scale unstructured volume data with mixed cell types is to dissolve a bottleneck caused by visibility sorting of unstructured cells. In this paper, we implement an...
详细信息
ISBN:
(纸本)9781509057382
The primary challenge in the visualization of a large-scale unstructured volume data with mixed cell types is to dissolve a bottleneck caused by visibility sorting of unstructured cells. In this paper, we implement an interactive particle-based rendering method to solve this complex visualization problem. This technique uses opaque particles as the proxy geometry of the mixed unstructured cells so that the visibility sorting process is not needed. This characteristic makes the rendering of mixed-cell unstructured volume efficient. We also construct a resizing function to adjust the particle radius depending on the assigned transfer function so that the transfer function can be adjusted in real time. Furthermore, we develop a time-varying level-of-detail (LOD) rendering to efficiently handle the large-scale time-varying data. This LOD rendering can provide highspeed rendering for animation rendering and high-quality rendering when the animation is stopped at any time step of interest. These features facilitate the detailed analysis of the temporal features of the data.
In recent years online display advertising has grown at a rapid pace. Genome from Yahoo! is the big data buying solution for online display advertising. The goal of our platform is to identify the best opportunity to ...
详细信息
Three-dimensional thermal analysis for the stator end-core of large turbine-generators, which has not been completely solved yet, is developed based on flow velocity measurements using a visual model and on the result...
详细信息
Three-dimensional thermal analysis for the stator end-core of large turbine-generators, which has not been completely solved yet, is developed based on flow velocity measurements using a visual model and on the results of core loss analysis, compared with test data.
The generation of short pulses of ion beams through the interaction of an intense laser with a plasma sheath offers the possibility of compact and cheaper ion sources for many applications--from fast ignition and radi...
详细信息
The generation of short pulses of ion beams through the interaction of an intense laser with a plasma sheath offers the possibility of compact and cheaper ion sources for many applications--from fast ignition and radiography of dense targets to hadron therapy and injection into conventional accelerators. To enable the efficient analysis of large-scale, high-fidelity particle accelerator simulations using the Warp simulation suite, the authors introduce the Warp In situ visualization Toolkit (WarpIV). WarpIV integrates state-of-the-art in situ visualization and analysis using VisIt with Warp, supports management and control of complex in situ visualization and analysis workflows, and implements integrated analytics to facilitate query- and feature-based data analytics and efficient large-scale dataanalysis. WarpIV enables for the first time distributed parallel, in situ visualization of the full simulation data using high-performance compute resources as the data is being generated by Warp. The authors describe the application of WarpIV to study and compare large 2D and 3D ion accelerator simulations, demonstrating significant differences in the acceleration process in 2D and 3D simulations. WarpIV is available to the public via https://***/berkeleylab/warpiv. The Warp In situ visualization Toolkit (WarpIV) supports large-scale, parallel, in situ visualization and analysis and facilitates query- and feature-based analytics, enabling for the first time high-performance analysis of large-scale, high-fidelity particle accelerator simulations while the data is being generated by the Warp simulation suite. This supplemental material https://***/extra/*** provides more details regarding the memory profiling and optimization and the Yee grid recentering optimization results discussed in the main article. [ABSTRACT FROM PUBLISHER]
Existing dataanalysis and visualization algorithms are used in a wide range of simulations that strive to support an increasing number of runtime systems. The BabelFlow framework has been designed to address this sit...
详细信息
ISBN:
(纸本)9781665432832
Existing dataanalysis and visualization algorithms are used in a wide range of simulations that strive to support an increasing number of runtime systems. The BabelFlow framework has been designed to address this situation by providing users with a simple interface to implement analysis algorithms as dataflow graphs portable across different runtimes. The limitation in BabelFlow, however, is that the graphs are not easily reusable. Plugging them into existing in situ workflows and constructing more complex graphs is difficult. In this paper, we introduce LegoFlow, an extension to BabelFlow that addresses these challenges. Specifically, we integrate LegoFlow into Ascent, a flyweight framework for large scale in situ analytics, and provide a graph composability mechanism. This mechanism is an intuitive approach to link an arbitrary number of graphs together to create more complex patterns, as well as avoid costly reimplementations for minor modifications. Without sacrificing portability, LegoFlow introduces complete flexibility that maximizes the productivity of in situ analytics workflows. Furthermore, we demonstrate a complete LULESH simulation with LegoFlow-based in situ visualization running on top of Charm++. It is a novel approach for in situ analytics, whereby the asynchronous tasking runtime allows routines for computation and analysis to overlap. Finally, we evaluate a number of LegoFlow-based filters and extracts in Ascent, as well as the scaling behavior of a LegoFlow graph for Radix-k based image compositing.
Tracking the temporal evolution of features in time-varying data remains a combinatorially challenging problem. A recent method models event detection as a maximum-weight independent set problem on a graph representat...
详细信息
ISBN:
(纸本)9781728126050
Tracking the temporal evolution of features in time-varying data remains a combinatorially challenging problem. A recent method models event detection as a maximum-weight independent set problem on a graph representation of all possible explanations [35]. However, optimally solving this problem is NP-hard in the general case. Following the approach by Schnorr et al., we propose a new algorithm for event detection. Our algorithm exploits the model-specific structure of the independent set problem. Specifically, we show how to traverse potential explanations in such a way that a greedy assignment provides reliably good results. We demonstrate the effectiveness of our approach on synthetic and simulation data sets, the former of which include ground-truth tracking information which enable a quantitative evaluation. Our results are within 1% of the theoretical optimum and comparable to an approximate solution provided by a state-of-the-art optimization package. At the same time, our algorithm is significantly faster.
We present a new parallel algorithm for probabilistic graphical model optimization. The algorithm relies on data-parallel primitives (DPPs), which provide portable performance over hardware architecture. We evaluate r...
详细信息
ISBN:
(纸本)9781538668733
We present a new parallel algorithm for probabilistic graphical model optimization. The algorithm relies on data-parallel primitives (DPPs), which provide portable performance over hardware architecture. We evaluate results on CPUs and GPUs for an image segmentation problem. Compared to a serial baseline, we observe runtime speedups of up to 13X (CPU) and 44X (GPU). We also compare our performance to a reference, OpenMP-based algorithm, and find speedups of up to 7X (CPU).
暂无评论