From mesh generation to scientific visualization: An end-to-end approach to parallel supercomputing

Tiankai Tu, Hongfeng Yu, Leonardo Ramirez-Guzman, Jacobo Bielak, Omar Ghattas, Kwan-Liu Ma, David R. O'Hallaron

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

92 Citations (Scopus)

Abstract

Parallel supercomputing has traditionally focused on the inner kernel of scientific simulations: the solver. The front and back ends of the simulation pipeline - -problem description and interpretation of the output - -have taken a back seat to the solver when it comes to attention paid to scalability and performance, and are often relegated to offline, sequential computation. As the largest simulations move beyond the realm of the terascale and into the petascale, this decomposition in tasks and platforms becomes increasingly untenable. We propose an end-to-end approach in which all simulation components - -meshing, partitioning, solver, and visualization - -are tightly coupled and execute in parallel with shared data structures and no intermediate I/O. We present our implementation of this new approach in the context of octree-based finite element simulation of earthquake ground motion. Performance evaluation on up to 2048 processors demonstrates the ability of the end-toend approach to overcome the scalability bottlenecks of the traditional approach.

Original languageEnglish (US)
Title of host publicationProceedings of the 2006 ACM/IEEE Conference on Supercomputing, SC'06
DOIs
StatePublished - Dec 1 2006

Fingerprint

Mesh generation
Data visualization
Scalability
Seats
Data structures
Earthquakes
Visualization
Pipelines
Decomposition

ASJC Scopus subject areas

  • Computer Science(all)

Cite this

Tu, T., Yu, H., Ramirez-Guzman, L., Bielak, J., Ghattas, O., Ma, K-L., & O'Hallaron, D. R. (2006). From mesh generation to scientific visualization: An end-to-end approach to parallel supercomputing. In Proceedings of the 2006 ACM/IEEE Conference on Supercomputing, SC'06 [1188551] https://doi.org/10.1145/1188455.1188551

From mesh generation to scientific visualization : An end-to-end approach to parallel supercomputing. / Tu, Tiankai; Yu, Hongfeng; Ramirez-Guzman, Leonardo; Bielak, Jacobo; Ghattas, Omar; Ma, Kwan-Liu; O'Hallaron, David R.

Proceedings of the 2006 ACM/IEEE Conference on Supercomputing, SC'06. 2006. 1188551.

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

Tu, T, Yu, H, Ramirez-Guzman, L, Bielak, J, Ghattas, O, Ma, K-L & O'Hallaron, DR 2006, From mesh generation to scientific visualization: An end-to-end approach to parallel supercomputing. in Proceedings of the 2006 ACM/IEEE Conference on Supercomputing, SC'06., 1188551. https://doi.org/10.1145/1188455.1188551
Tu T, Yu H, Ramirez-Guzman L, Bielak J, Ghattas O, Ma K-L et al. From mesh generation to scientific visualization: An end-to-end approach to parallel supercomputing. In Proceedings of the 2006 ACM/IEEE Conference on Supercomputing, SC'06. 2006. 1188551 https://doi.org/10.1145/1188455.1188551
Tu, Tiankai ; Yu, Hongfeng ; Ramirez-Guzman, Leonardo ; Bielak, Jacobo ; Ghattas, Omar ; Ma, Kwan-Liu ; O'Hallaron, David R. / From mesh generation to scientific visualization : An end-to-end approach to parallel supercomputing. Proceedings of the 2006 ACM/IEEE Conference on Supercomputing, SC'06. 2006.
@inproceedings{4e5428415f124a88b262853bc4be70ae,
title = "From mesh generation to scientific visualization: An end-to-end approach to parallel supercomputing",
abstract = "Parallel supercomputing has traditionally focused on the inner kernel of scientific simulations: the solver. The front and back ends of the simulation pipeline - -problem description and interpretation of the output - -have taken a back seat to the solver when it comes to attention paid to scalability and performance, and are often relegated to offline, sequential computation. As the largest simulations move beyond the realm of the terascale and into the petascale, this decomposition in tasks and platforms becomes increasingly untenable. We propose an end-to-end approach in which all simulation components - -meshing, partitioning, solver, and visualization - -are tightly coupled and execute in parallel with shared data structures and no intermediate I/O. We present our implementation of this new approach in the context of octree-based finite element simulation of earthquake ground motion. Performance evaluation on up to 2048 processors demonstrates the ability of the end-toend approach to overcome the scalability bottlenecks of the traditional approach.",
author = "Tiankai Tu and Hongfeng Yu and Leonardo Ramirez-Guzman and Jacobo Bielak and Omar Ghattas and Kwan-Liu Ma and O'Hallaron, {David R.}",
year = "2006",
month = "12",
day = "1",
doi = "10.1145/1188455.1188551",
language = "English (US)",
isbn = "0769527000",
booktitle = "Proceedings of the 2006 ACM/IEEE Conference on Supercomputing, SC'06",

}

TY - GEN

T1 - From mesh generation to scientific visualization

T2 - An end-to-end approach to parallel supercomputing

AU - Tu, Tiankai

AU - Yu, Hongfeng

AU - Ramirez-Guzman, Leonardo

AU - Bielak, Jacobo

AU - Ghattas, Omar

AU - Ma, Kwan-Liu

AU - O'Hallaron, David R.

PY - 2006/12/1

Y1 - 2006/12/1

N2 - Parallel supercomputing has traditionally focused on the inner kernel of scientific simulations: the solver. The front and back ends of the simulation pipeline - -problem description and interpretation of the output - -have taken a back seat to the solver when it comes to attention paid to scalability and performance, and are often relegated to offline, sequential computation. As the largest simulations move beyond the realm of the terascale and into the petascale, this decomposition in tasks and platforms becomes increasingly untenable. We propose an end-to-end approach in which all simulation components - -meshing, partitioning, solver, and visualization - -are tightly coupled and execute in parallel with shared data structures and no intermediate I/O. We present our implementation of this new approach in the context of octree-based finite element simulation of earthquake ground motion. Performance evaluation on up to 2048 processors demonstrates the ability of the end-toend approach to overcome the scalability bottlenecks of the traditional approach.

AB - Parallel supercomputing has traditionally focused on the inner kernel of scientific simulations: the solver. The front and back ends of the simulation pipeline - -problem description and interpretation of the output - -have taken a back seat to the solver when it comes to attention paid to scalability and performance, and are often relegated to offline, sequential computation. As the largest simulations move beyond the realm of the terascale and into the petascale, this decomposition in tasks and platforms becomes increasingly untenable. We propose an end-to-end approach in which all simulation components - -meshing, partitioning, solver, and visualization - -are tightly coupled and execute in parallel with shared data structures and no intermediate I/O. We present our implementation of this new approach in the context of octree-based finite element simulation of earthquake ground motion. Performance evaluation on up to 2048 processors demonstrates the ability of the end-toend approach to overcome the scalability bottlenecks of the traditional approach.

UR - http://www.scopus.com/inward/record.url?scp=34548281709&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34548281709&partnerID=8YFLogxK

U2 - 10.1145/1188455.1188551

DO - 10.1145/1188455.1188551

M3 - Conference contribution

AN - SCOPUS:34548281709

SN - 0769527000

SN - 9780769527000

BT - Proceedings of the 2006 ACM/IEEE Conference on Supercomputing, SC'06

ER -