A Lifeline-Based Approach for Work Requesting and Parallel Particle Advection

Roba Binyahib; David Pugmire; Boyana Norris; Hank Childs

doi:10.1109/LDAV48142.2019.8944355

A Lifeline-Based Approach for Work Requesting and Parallel Particle Advection

Roba Binyahib, David Pugmire, Boyana Norris, Hank Childs

Visualization

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

8 Scopus citations

Abstract

Particle advection, a fundamental building block for many flow visualization algorithms, is very difficult to parallelize efficiently. That said, work requesting is a promising technique to improve parallel performance for particle advection. With this work, we introduce a new work requesting-based method which uses the Lifeline scheduling method. To evaluate the impact of this new algorithm, we ran 92 experiments, running at concurrencies as high as 8192 cores, data sets as large as 17 billion cells, and as many as 16 million particles, comparing against other work requesting scheduling methods. Overall, our results show that Lifeline has significantly less idle time than other approaches, since it reduces the number of failed attempts to request work.

Original language	English
Title of host publication	2019 IEEE 9th Symposium on Large Data Analysis and Visualization, LDAV 2019
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	52-61
Number of pages	10
ISBN (Electronic)	9781728126050
DOIs	https://doi.org/10.1109/LDAV48142.2019.8944355
State	Published - Oct 2019
Event	9th IEEE Symposium on Large-Scale Data Analysis and Visualization, LDAV 2019 - Vancouver, Canada Duration: Oct 21 2019 → …

Publication series

Name	2019 IEEE 9th Symposium on Large Data Analysis and Visualization, LDAV 2019

Conference

Conference	9th IEEE Symposium on Large-Scale Data Analysis and Visualization, LDAV 2019
Country/Territory	Canada
City	Vancouver
Period	10/21/19 → …

Funding

This research was supported by the Exascale Computing Project (17-SC-20-SC), a collaborative effort of the U.S. Department of Energy Office of Science and the National Nuclear Security Administration. This research was supported by the Scientific Discovery through Advanced Computing (SciDAC) program of the U.S. Department of Energy. This research used resources of the National Energy Research Scientific Computing Center (NERSC), a U.S. Department of Energy Office of Science User Facility operated under Contract No. DE-AC02-05CH11231.

Keywords

Lifeline-Based scheduling
Parallel particle advection
Visualization
Work requesting

Access to Document

10.1109/LDAV48142.2019.8944355

Cite this

Binyahib, R., Pugmire, D., Norris, B., & Childs, H. (2019). A Lifeline-Based Approach for Work Requesting and Parallel Particle Advection. In 2019 IEEE 9th Symposium on Large Data Analysis and Visualization, LDAV 2019 (pp. 52-61). Article 8944355 (2019 IEEE 9th Symposium on Large Data Analysis and Visualization, LDAV 2019). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/LDAV48142.2019.8944355

@inproceedings{ad2b0b324b6347f8a9421bb12a6b8e44,

title = "A Lifeline-Based Approach for Work Requesting and Parallel Particle Advection",

abstract = "Particle advection, a fundamental building block for many flow visualization algorithms, is very difficult to parallelize efficiently. That said, work requesting is a promising technique to improve parallel performance for particle advection. With this work, we introduce a new work requesting-based method which uses the Lifeline scheduling method. To evaluate the impact of this new algorithm, we ran 92 experiments, running at concurrencies as high as 8192 cores, data sets as large as 17 billion cells, and as many as 16 million particles, comparing against other work requesting scheduling methods. Overall, our results show that Lifeline has significantly less idle time than other approaches, since it reduces the number of failed attempts to request work.",

keywords = "Lifeline-Based scheduling, Parallel particle advection, Visualization, Work requesting",

author = "Roba Binyahib and David Pugmire and Boyana Norris and Hank Childs",

note = "Publisher Copyright: {\textcopyright} 2019 IEEE.; 9th IEEE Symposium on Large-Scale Data Analysis and Visualization, LDAV 2019 ; Conference date: 21-10-2019",

year = "2019",

month = oct,

doi = "10.1109/LDAV48142.2019.8944355",

language = "English",

series = "2019 IEEE 9th Symposium on Large Data Analysis and Visualization, LDAV 2019",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "52--61",

booktitle = "2019 IEEE 9th Symposium on Large Data Analysis and Visualization, LDAV 2019",

}

Binyahib, R, Pugmire, D, Norris, B & Childs, H 2019, A Lifeline-Based Approach for Work Requesting and Parallel Particle Advection. in 2019 IEEE 9th Symposium on Large Data Analysis and Visualization, LDAV 2019., 8944355, 2019 IEEE 9th Symposium on Large Data Analysis and Visualization, LDAV 2019, Institute of Electrical and Electronics Engineers Inc., pp. 52-61, 9th IEEE Symposium on Large-Scale Data Analysis and Visualization, LDAV 2019, Vancouver, Canada, 10/21/19. https://doi.org/10.1109/LDAV48142.2019.8944355

A Lifeline-Based Approach for Work Requesting and Parallel Particle Advection. / Binyahib, Roba; Pugmire, David; Norris, Boyana et al.
2019 IEEE 9th Symposium on Large Data Analysis and Visualization, LDAV 2019. Institute of Electrical and Electronics Engineers Inc., 2019. p. 52-61 8944355 (2019 IEEE 9th Symposium on Large Data Analysis and Visualization, LDAV 2019).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - A Lifeline-Based Approach for Work Requesting and Parallel Particle Advection

AU - Binyahib, Roba

AU - Pugmire, David

AU - Norris, Boyana

AU - Childs, Hank

PY - 2019/10

Y1 - 2019/10

N2 - Particle advection, a fundamental building block for many flow visualization algorithms, is very difficult to parallelize efficiently. That said, work requesting is a promising technique to improve parallel performance for particle advection. With this work, we introduce a new work requesting-based method which uses the Lifeline scheduling method. To evaluate the impact of this new algorithm, we ran 92 experiments, running at concurrencies as high as 8192 cores, data sets as large as 17 billion cells, and as many as 16 million particles, comparing against other work requesting scheduling methods. Overall, our results show that Lifeline has significantly less idle time than other approaches, since it reduces the number of failed attempts to request work.

AB - Particle advection, a fundamental building block for many flow visualization algorithms, is very difficult to parallelize efficiently. That said, work requesting is a promising technique to improve parallel performance for particle advection. With this work, we introduce a new work requesting-based method which uses the Lifeline scheduling method. To evaluate the impact of this new algorithm, we ran 92 experiments, running at concurrencies as high as 8192 cores, data sets as large as 17 billion cells, and as many as 16 million particles, comparing against other work requesting scheduling methods. Overall, our results show that Lifeline has significantly less idle time than other approaches, since it reduces the number of failed attempts to request work.

KW - Lifeline-Based scheduling

KW - Parallel particle advection

KW - Visualization

KW - Work requesting

UR - http://www.scopus.com/inward/record.url?scp=85078105176&partnerID=8YFLogxK

U2 - 10.1109/LDAV48142.2019.8944355

DO - 10.1109/LDAV48142.2019.8944355

M3 - Conference contribution

AN - SCOPUS:85078105176

T3 - 2019 IEEE 9th Symposium on Large Data Analysis and Visualization, LDAV 2019

SP - 52

EP - 61

BT - 2019 IEEE 9th Symposium on Large Data Analysis and Visualization, LDAV 2019

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 9th IEEE Symposium on Large-Scale Data Analysis and Visualization, LDAV 2019

Y2 - 21 October 2019

ER -

Binyahib R, Pugmire D, Norris B, Childs H. A Lifeline-Based Approach for Work Requesting and Parallel Particle Advection. In 2019 IEEE 9th Symposium on Large Data Analysis and Visualization, LDAV 2019. Institute of Electrical and Electronics Engineers Inc. 2019. p. 52-61. 8944355. (2019 IEEE 9th Symposium on Large Data Analysis and Visualization, LDAV 2019). doi: 10.1109/LDAV48142.2019.8944355

A Lifeline-Based Approach for Work Requesting and Parallel Particle Advection

Abstract

Publication series

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this