Large volume visualization of compressed time-dependent datasets on GPU clusters

Strengert, Magnus; Magallón, Marcelo; Weiskopf, Daniel; Guthe, Stefan; Ertl, Thomas

doi:10.1016/j.parco.2005.02.006

Cited by 24 publications

(23 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Using multi-GPU systems, including GPU clusters, is gaining popularity in scientific computing [5,8,9,11,20,23]. In general, these works demonstrate that such platforms can be beneficial in terms of performance, power, and price.…”

Section: Related Workmentioning

confidence: 99%

Exploring Fine-Grained Task-Based Execution on Multi-GPU Systems

Chen

Villa

Gao

2011

2011 IEEE International Conference on Cluster Computing

View full text Add to dashboard Cite

Using multi-GPU systems, including GPU clusters, is gaining popularity in scientific computing. However, when using multiple GPUs concurrently, the conventional data parallel GPU programming paradigms, e.g., CUDA, cannot satisfactorily address certain issues, such as load balancing, GPU resource utilization, overlapping finegrained computation with communication, etc. In this paper, we present a fine-grained task-based execution framework for multi-GPU systems. By scheduling finer-grained tasks than what is supported in the conventional CUDA programming method among multiple GPUs, and allowing concurrent task execution on a single GPU, our framework provides means for solving the above issues and efficiently utilizing multi-GPU systems. Experiments with a molecular dynamics application show that, for nonuniform distributed workload, the solutions based on our framework achieve good load balance, and considerable performance improvement over other solutions based on the standard CUDA programming methodologies.

show abstract

Section: Related Workmentioning

confidence: 99%

Exploring Fine-Grained Task-Based Execution on Multi-GPU Systems

Chen

Villa

Gao

2011

2011 IEEE International Conference on Cluster Computing

View full text Add to dashboard Cite

show abstract

“…If the data amount exceeds the memory of the CPU or GPU, several techniques can be employed, including compressed or packed representations of the data [29], decomposition techniques, multi-resolution schemes [20,53,54], or out-of-core techniques [18]. Recent research combined bricking and decomposition with a hierarchical data structure.…”

Section: Software Techniques Coping With Large Datamentioning

confidence: 99%

“…Parallel GPU-based programming on a single node with one GPU or multiple GPUs using programming languages for the massive parallel cores on the graphic card [53][54][55]62]. With advances in GPU architecture, several algorithms have reached higher efficiency by transferring the program from CPU to GPU.…”

Section: Software Techniques Coping With Large Datamentioning

confidence: 99%

Challenges of medical image processing

et al. 2010

View full text Add to dashboard Cite

In todays health care, imaging plays an important role throughout the entire clinical process from diagnostics and treatment planning to surgical procedures and follow up studies. Since most imaging modalities have gone directly digital, with continually increasing resolution, medical image processing has to face the challenges arising from large data volumes. In this paper, we discuss Kilo-to Terabyte challenges regarding (i) medical image management and image data mining, (ii) bioimaging, (iii) virtual reality in medical visualizations and (iv) neuroimaging. Due to the increasing amount of data, image processing and visualization algorithms have to be adjusted. Scalable algorithms and advanced parallelization techniques using graphical processing units have been developed. They are summarized in this paper. While such techniques are coping with the Kilo-to Terabyte challenge, the Petabyte level is already looming on the horizon. For this reason, medical image processing remains a vital field of research.

show abstract

“…Parallel visualization is also very often used to address the issues of large data processing [2,4,15,22]. The overwhelming majority of such cases employ a divideand-conquer strategy that involves the subdivision of the large volume dataset into parts small enough to be processed on a single processor, successive, or simultaneous processing (e.g., visualization) of these parts, and the merging of outcomes to produce the final result.…”

Section: Introductionmentioning

confidence: 99%

“…There are also approaches that use wavelet compression to reduce the size of the volume data, and thus make the retention of the entire dataset in memory possible or speed up the transmission of data over the network in parallel visualization [9,22].…”

Section: Introductionmentioning

confidence: 99%

Enabling the interactive display of large medical volume datasets by multiresolution bricking

et al. 2009

View full text Add to dashboard Cite

In this paper, we present an approach to interactive out-of-core volume data exploration that has been developed to augment the existing capabilities of the LhpBuilder software, a core component of the European project LHDL (http://www.biomedtown.org/biomed_town/lhdl). The requirements relate to importing, accessing, visualizing and extracting a part of a very large volume dataset by interactive visual exploration. Such datasets contain billions of voxels and, therefore, several gigabytes are required just to store them, which quickly surpass the virtual address limit of current 32-bit PC platforms. We have implemented a hierarchical, bricked, partition-based, out-of-core strategy to balance the usage of main and external memories. A new indexing scheme is introduced, which permits the use of a multiresolution bricked volume layout with minimum overhead and also supports fast data compression. Using the hierarchy constructed in a pre-processing step, we generate a coarse approximation that provides a preview using direct volume visualization for large-scale datasets. A user can interactively explore the dataset by specifying a region of interest (ROI), which further generates a much more accurate data representation inside the ROI. If even more precise accuracy is needed inside the ROI, nested ROIs are used. The software has been constructed using the Multimod Application Framework, a VTK-based system; however, the approach can be adopted for the other systems in a straightforward way. Experimental results show that the user can interactively explore large volume datasets such as the Visible Human Male/Female (with

show abstract

Large volume visualization of compressed time-dependent datasets on GPU clusters

Cited by 24 publications

References 12 publications

Exploring Fine-Grained Task-Based Execution on Multi-GPU Systems

Exploring Fine-Grained Task-Based Execution on Multi-GPU Systems

Challenges of medical image processing

Enabling the interactive display of large medical volume datasets by multiresolution bricking

Contact Info

Product

Resources

About