Multi-step image compositing for massively parallel rendering

Jorji Nonaka, Kenji Ono, Masahiro Fujita

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

High performance visualization has played an important role in computer-aided scientific discovery and has become an indispensable tool for computational scientists. Sort-last parallel rendering is a proven approach for visual data analytics by extracting meaningful information from huge data sets generated from large scale scientific computing. Image compositing is the last stage of sort-last parallel rendering pipeline and works by combining the images generated by the rendering nodes to generate the final image. Since it requires interprocess communication among the entire nodes, it usually dominates the total cost of the parallel rendering process. In current high-end massively parallel HPC systems, where tens or even hundreds of thousands of nodes can be involved, performance degradation is inevitable even using theoretically scalable image compositing algortithms such as the well-known Binary-Swap method. To minimize this undesirable performance degradation, we propose the multi-step image compositing method, where the image compositing nodes are divided into smaller groups and the entire process is performed in several steps. We evaluated the proposed image compositing method on RIKEN K computer, which is a massively parallel HPC system, and we obtained encouraging results showing the effectiveness of this method in a large-scale image compositing environment. We also foresee a great potential of this method to meet the large-scale image compositing demands brought about by the rapid increase in processor counts of current and next-generation HPC systems.

Original languageEnglish
Title of host publicationProceedings of the 2014 International Conference on High Performance Computing and Simulation, HPCS 2014
EditorsWaleed Smari, Vesna Zeljkovic
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages627-634
Number of pages8
ISBN (Electronic)9781479953127
DOIs
Publication statusPublished - Sep 18 2014
Event2014 International Conference on High Performance Computing and Simulation, HPCS 2014 - Bologna, Italy
Duration: Jul 21 2014Jul 25 2014

Publication series

NameProceedings of the 2014 International Conference on High Performance Computing and Simulation, HPCS 2014

Other

Other2014 International Conference on High Performance Computing and Simulation, HPCS 2014
CountryItaly
CityBologna
Period7/21/147/25/14

Fingerprint

Rendering
Natural sciences computing
Degradation
Visualization
Pipelines
Communication
Parallel Systems
Vertex of a graph
Costs
Sort
Entire
Rendering (computer graphics)
Scientific Computing
Swap
Count
High Performance
Binary
Minimise

All Science Journal Classification (ASJC) codes

  • Computer Networks and Communications
  • Software
  • Modelling and Simulation

Cite this

Nonaka, J., Ono, K., & Fujita, M. (2014). Multi-step image compositing for massively parallel rendering. In W. Smari, & V. Zeljkovic (Eds.), Proceedings of the 2014 International Conference on High Performance Computing and Simulation, HPCS 2014 (pp. 627-634). [6903746] (Proceedings of the 2014 International Conference on High Performance Computing and Simulation, HPCS 2014). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/HPCSim.2014.6903746

Multi-step image compositing for massively parallel rendering. / Nonaka, Jorji; Ono, Kenji; Fujita, Masahiro.

Proceedings of the 2014 International Conference on High Performance Computing and Simulation, HPCS 2014. ed. / Waleed Smari; Vesna Zeljkovic. Institute of Electrical and Electronics Engineers Inc., 2014. p. 627-634 6903746 (Proceedings of the 2014 International Conference on High Performance Computing and Simulation, HPCS 2014).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Nonaka, J, Ono, K & Fujita, M 2014, Multi-step image compositing for massively parallel rendering. in W Smari & V Zeljkovic (eds), Proceedings of the 2014 International Conference on High Performance Computing and Simulation, HPCS 2014., 6903746, Proceedings of the 2014 International Conference on High Performance Computing and Simulation, HPCS 2014, Institute of Electrical and Electronics Engineers Inc., pp. 627-634, 2014 International Conference on High Performance Computing and Simulation, HPCS 2014, Bologna, Italy, 7/21/14. https://doi.org/10.1109/HPCSim.2014.6903746
Nonaka J, Ono K, Fujita M. Multi-step image compositing for massively parallel rendering. In Smari W, Zeljkovic V, editors, Proceedings of the 2014 International Conference on High Performance Computing and Simulation, HPCS 2014. Institute of Electrical and Electronics Engineers Inc. 2014. p. 627-634. 6903746. (Proceedings of the 2014 International Conference on High Performance Computing and Simulation, HPCS 2014). https://doi.org/10.1109/HPCSim.2014.6903746
Nonaka, Jorji ; Ono, Kenji ; Fujita, Masahiro. / Multi-step image compositing for massively parallel rendering. Proceedings of the 2014 International Conference on High Performance Computing and Simulation, HPCS 2014. editor / Waleed Smari ; Vesna Zeljkovic. Institute of Electrical and Electronics Engineers Inc., 2014. pp. 627-634 (Proceedings of the 2014 International Conference on High Performance Computing and Simulation, HPCS 2014).
@inproceedings{3f69338c68e64d0bba89e49390c7608b,
title = "Multi-step image compositing for massively parallel rendering",
abstract = "High performance visualization has played an important role in computer-aided scientific discovery and has become an indispensable tool for computational scientists. Sort-last parallel rendering is a proven approach for visual data analytics by extracting meaningful information from huge data sets generated from large scale scientific computing. Image compositing is the last stage of sort-last parallel rendering pipeline and works by combining the images generated by the rendering nodes to generate the final image. Since it requires interprocess communication among the entire nodes, it usually dominates the total cost of the parallel rendering process. In current high-end massively parallel HPC systems, where tens or even hundreds of thousands of nodes can be involved, performance degradation is inevitable even using theoretically scalable image compositing algortithms such as the well-known Binary-Swap method. To minimize this undesirable performance degradation, we propose the multi-step image compositing method, where the image compositing nodes are divided into smaller groups and the entire process is performed in several steps. We evaluated the proposed image compositing method on RIKEN K computer, which is a massively parallel HPC system, and we obtained encouraging results showing the effectiveness of this method in a large-scale image compositing environment. We also foresee a great potential of this method to meet the large-scale image compositing demands brought about by the rapid increase in processor counts of current and next-generation HPC systems.",
author = "Jorji Nonaka and Kenji Ono and Masahiro Fujita",
year = "2014",
month = "9",
day = "18",
doi = "10.1109/HPCSim.2014.6903746",
language = "English",
series = "Proceedings of the 2014 International Conference on High Performance Computing and Simulation, HPCS 2014",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "627--634",
editor = "Waleed Smari and Vesna Zeljkovic",
booktitle = "Proceedings of the 2014 International Conference on High Performance Computing and Simulation, HPCS 2014",
address = "United States",

}

TY - GEN

T1 - Multi-step image compositing for massively parallel rendering

AU - Nonaka, Jorji

AU - Ono, Kenji

AU - Fujita, Masahiro

PY - 2014/9/18

Y1 - 2014/9/18

N2 - High performance visualization has played an important role in computer-aided scientific discovery and has become an indispensable tool for computational scientists. Sort-last parallel rendering is a proven approach for visual data analytics by extracting meaningful information from huge data sets generated from large scale scientific computing. Image compositing is the last stage of sort-last parallel rendering pipeline and works by combining the images generated by the rendering nodes to generate the final image. Since it requires interprocess communication among the entire nodes, it usually dominates the total cost of the parallel rendering process. In current high-end massively parallel HPC systems, where tens or even hundreds of thousands of nodes can be involved, performance degradation is inevitable even using theoretically scalable image compositing algortithms such as the well-known Binary-Swap method. To minimize this undesirable performance degradation, we propose the multi-step image compositing method, where the image compositing nodes are divided into smaller groups and the entire process is performed in several steps. We evaluated the proposed image compositing method on RIKEN K computer, which is a massively parallel HPC system, and we obtained encouraging results showing the effectiveness of this method in a large-scale image compositing environment. We also foresee a great potential of this method to meet the large-scale image compositing demands brought about by the rapid increase in processor counts of current and next-generation HPC systems.

AB - High performance visualization has played an important role in computer-aided scientific discovery and has become an indispensable tool for computational scientists. Sort-last parallel rendering is a proven approach for visual data analytics by extracting meaningful information from huge data sets generated from large scale scientific computing. Image compositing is the last stage of sort-last parallel rendering pipeline and works by combining the images generated by the rendering nodes to generate the final image. Since it requires interprocess communication among the entire nodes, it usually dominates the total cost of the parallel rendering process. In current high-end massively parallel HPC systems, where tens or even hundreds of thousands of nodes can be involved, performance degradation is inevitable even using theoretically scalable image compositing algortithms such as the well-known Binary-Swap method. To minimize this undesirable performance degradation, we propose the multi-step image compositing method, where the image compositing nodes are divided into smaller groups and the entire process is performed in several steps. We evaluated the proposed image compositing method on RIKEN K computer, which is a massively parallel HPC system, and we obtained encouraging results showing the effectiveness of this method in a large-scale image compositing environment. We also foresee a great potential of this method to meet the large-scale image compositing demands brought about by the rapid increase in processor counts of current and next-generation HPC systems.

UR - http://www.scopus.com/inward/record.url?scp=84908650653&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84908650653&partnerID=8YFLogxK

U2 - 10.1109/HPCSim.2014.6903746

DO - 10.1109/HPCSim.2014.6903746

M3 - Conference contribution

AN - SCOPUS:84908650653

T3 - Proceedings of the 2014 International Conference on High Performance Computing and Simulation, HPCS 2014

SP - 627

EP - 634

BT - Proceedings of the 2014 International Conference on High Performance Computing and Simulation, HPCS 2014

A2 - Smari, Waleed

A2 - Zeljkovic, Vesna

PB - Institute of Electrical and Electronics Engineers Inc.

ER -