Multi-Step Image Composition Approach for Sort-Last Massively Parallel Rendering

Jorji Nonaka, Masahiro Fujita, Kenji Ono

Research output: Contribution to journalArticle

Abstract

Large-scale numerical simulations on modern leading-edge supercomputers have been continuously generating tremendous amount of data. <i>In-Situ Visualization</i> is widely recognized as the most rational way for analysis and mining of such large data sets by the use of sort-last parallel visualization. However, sort-last method requires communication intensive final image composition and can suffer from scalability problem on massively parallel rendering and compositing environments. In this paper, we present the <i>Multi-Step Image Composition</i> approach to achieve scalability by minimizing undesirable performance degradation on such massively parallel rendering environments. We verified the effectiveness of this proposed approach on K computer, installed at RIKEN AICS, and achieved a speedup of 1.8× to 7.8× using 32,768 composition nodes and different image sizes. We foresee a great potential of this method to meet the even larger image composition demands brought about by the rapid increase in the number of processing elements on modern HPC systems.
Original languageEnglish
Pages (from-to)108-125
Number of pages18
JournalJASSE
Volume2
Issue number1
DOIs
Publication statusPublished - 2015

Fingerprint

Chemical analysis
Scalability
Visualization
Supercomputers
Degradation
Communication
Computer simulation
Processing

Cite this

Multi-Step Image Composition Approach for Sort-Last Massively Parallel Rendering. / Nonaka, Jorji; Fujita, Masahiro; Ono, Kenji.

In: JASSE, Vol. 2, No. 1, 2015, p. 108-125.

Research output: Contribution to journalArticle

Nonaka, Jorji ; Fujita, Masahiro ; Ono, Kenji. / Multi-Step Image Composition Approach for Sort-Last Massively Parallel Rendering. In: JASSE. 2015 ; Vol. 2, No. 1. pp. 108-125.
@article{d9513d346ed14b1e88da591654ab402d,
title = "Multi-Step Image Composition Approach for Sort-Last Massively Parallel Rendering",
abstract = "Large-scale numerical simulations on modern leading-edge supercomputers have been continuously generating tremendous amount of data. In-Situ Visualization is widely recognized as the most rational way for analysis and mining of such large data sets by the use of sort-last parallel visualization. However, sort-last method requires communication intensive final image composition and can suffer from scalability problem on massively parallel rendering and compositing environments. In this paper, we present the Multi-Step Image Composition approach to achieve scalability by minimizing undesirable performance degradation on such massively parallel rendering environments. We verified the effectiveness of this proposed approach on K computer, installed at RIKEN AICS, and achieved a speedup of 1.8× to 7.8× using 32,768 composition nodes and different image sizes. We foresee a great potential of this method to meet the even larger image composition demands brought about by the rapid increase in the number of processing elements on modern HPC systems.",
author = "Jorji Nonaka and Masahiro Fujita and Kenji Ono",
year = "2015",
doi = "10.15748/jasse.2.108",
language = "English",
volume = "2",
pages = "108--125",
journal = "JASSE",
issn = "2188-5303",
publisher = "Japan Society for Simulation Technology, JSST",
number = "1",

}

TY - JOUR

T1 - Multi-Step Image Composition Approach for Sort-Last Massively Parallel Rendering

AU - Nonaka, Jorji

AU - Fujita, Masahiro

AU - Ono, Kenji

PY - 2015

Y1 - 2015

N2 - Large-scale numerical simulations on modern leading-edge supercomputers have been continuously generating tremendous amount of data. In-Situ Visualization is widely recognized as the most rational way for analysis and mining of such large data sets by the use of sort-last parallel visualization. However, sort-last method requires communication intensive final image composition and can suffer from scalability problem on massively parallel rendering and compositing environments. In this paper, we present the Multi-Step Image Composition approach to achieve scalability by minimizing undesirable performance degradation on such massively parallel rendering environments. We verified the effectiveness of this proposed approach on K computer, installed at RIKEN AICS, and achieved a speedup of 1.8× to 7.8× using 32,768 composition nodes and different image sizes. We foresee a great potential of this method to meet the even larger image composition demands brought about by the rapid increase in the number of processing elements on modern HPC systems.

AB - Large-scale numerical simulations on modern leading-edge supercomputers have been continuously generating tremendous amount of data. In-Situ Visualization is widely recognized as the most rational way for analysis and mining of such large data sets by the use of sort-last parallel visualization. However, sort-last method requires communication intensive final image composition and can suffer from scalability problem on massively parallel rendering and compositing environments. In this paper, we present the Multi-Step Image Composition approach to achieve scalability by minimizing undesirable performance degradation on such massively parallel rendering environments. We verified the effectiveness of this proposed approach on K computer, installed at RIKEN AICS, and achieved a speedup of 1.8× to 7.8× using 32,768 composition nodes and different image sizes. We foresee a great potential of this method to meet the even larger image composition demands brought about by the rapid increase in the number of processing elements on modern HPC systems.

U2 - 10.15748/jasse.2.108

DO - 10.15748/jasse.2.108

M3 - Article

VL - 2

SP - 108

EP - 125

JO - JASSE

JF - JASSE

SN - 2188-5303

IS - 1

ER -