UnSNAP: A Mini-App for Exploring the Performance of Deterministic Discrete Ordinates Transport on Unstructured Meshes

Tom Deakin; Simon McIntosh-Smith; Justin Lovegrove; Richard Smedley-Stevenson; Andrew Hagues

doi:10.1109/CLUSTER.2018.00076

UnSNAP: A Mini-App for Exploring the Performance of Deterministic Discrete Ordinates Transport on Unstructured Meshes

Tom Deakin, Simon McIntosh-Smith, Justin Lovegrove, Richard Smedley-Stevenson, Andrew Hagues

Research output: Chapter in Book/Report/Conference proceeding › Conference Contribution (Conference Proceeding)

3 Citations (Scopus)

360 Downloads (Pure)

Abstract

Solving the deterministic discrete ordinates neutral particle transport equation is a computationally expensive application. On an unstructured mesh, the discontinuous Galerkin finite element method is used for discretisation of the spatial domain. Additionally, an upwind dependency is applied forming wavefront sweeps across the spatial mesh for each iteration of the solve. We present a new mini-app, UnSNAP, which can be used to investigate the performance of arbitrarily high-order finite element unstructured transport on modern architectures. A new schedule appropriate for such architectures is presented. Finally, we show performance results for the mini-app on CPUs with high numbers of cores.

Original language	English
Title of host publication	Workshop of Representative Applications (WRAp) 18 at 2018 IEEE International Conference on Cluster Computing (CLUSTER 2018)
Publisher	IEEE Computer Society
Pages	598-606
Number of pages	9
ISBN (Electronic)	9781538683194
ISBN (Print)	9781538683200
DOIs	https://doi.org/10.1109/CLUSTER.2018.00076
Publication status	E-pub ahead of print - 1 Nov 2018

Publication series

Name
ISSN (Print)	1552-5244

Keywords

deterministic discrete ordinates transport
finite element
discontinuous Galerkin
sweep
unstructured mesh

Access to Document

10.1109/CLUSTER.2018.00076

Full-text PDF (accepted author manuscript)
This is the author accepted manuscript (AAM). The final published version (version of record) is available online via IEEE at https://ieeexplore.ieee.org/document/8514920 . Please refer to any applicable terms of use of the publisher.
Accepted author manuscript, 188 KB

Handle.net

Persistent link

Cite this

Deakin, T., McIntosh-Smith, S., Lovegrove, J., Smedley-Stevenson, R., & Hagues, A. (2018). UnSNAP: A Mini-App for Exploring the Performance of Deterministic Discrete Ordinates Transport on Unstructured Meshes. In Workshop of Representative Applications (WRAp) 18 at 2018 IEEE International Conference on Cluster Computing (CLUSTER 2018) (pp. 598-606). IEEE Computer Society. https://doi.org/10.1109/CLUSTER.2018.00076

@inproceedings{74079d3536a0495b9c5b6c780b5508f4,

title = "UnSNAP: A Mini-App for Exploring the Performance of Deterministic Discrete Ordinates Transport on Unstructured Meshes",

abstract = "Solving the deterministic discrete ordinates neutral particle transport equation is a computationally expensive application. On an unstructured mesh, the discontinuous Galerkin finite element method is used for discretisation of the spatial domain. Additionally, an upwind dependency is applied forming wavefront sweeps across the spatial mesh for each iteration of the solve. We present a new mini-app, UnSNAP, which can be used to investigate the performance of arbitrarily high-order finite element unstructured transport on modern architectures. A new schedule appropriate for such architectures is presented. Finally, we show performance results for the mini-app on CPUs with high numbers of cores.",

keywords = "deterministic discrete ordinates transport, finite element, discontinuous Galerkin, sweep, unstructured mesh",

author = "Tom Deakin and Simon McIntosh-Smith and Justin Lovegrove and Richard Smedley-Stevenson and Andrew Hagues",

year = "2018",

month = nov,

day = "1",

doi = "10.1109/CLUSTER.2018.00076",

language = "English",

isbn = "9781538683200",

publisher = "IEEE Computer Society",

pages = "598--606",

booktitle = "Workshop of Representative Applications (WRAp) 18 at 2018 IEEE International Conference on Cluster Computing (CLUSTER 2018)",

address = "United States",

}

Deakin, T , McIntosh-Smith, S, Lovegrove, J, Smedley-Stevenson, R & Hagues, A 2018, UnSNAP: A Mini-App for Exploring the Performance of Deterministic Discrete Ordinates Transport on Unstructured Meshes. in Workshop of Representative Applications (WRAp) 18 at 2018 IEEE International Conference on Cluster Computing (CLUSTER 2018). IEEE Computer Society, pp. 598-606. https://doi.org/10.1109/CLUSTER.2018.00076

UnSNAP: A Mini-App for Exploring the Performance of Deterministic Discrete Ordinates Transport on Unstructured Meshes. / Deakin, Tom ; McIntosh-Smith, Simon; Lovegrove, Justin et al.
Workshop of Representative Applications (WRAp) 18 at 2018 IEEE International Conference on Cluster Computing (CLUSTER 2018). IEEE Computer Society, 2018. p. 598-606.

Research output: Chapter in Book/Report/Conference proceeding › Conference Contribution (Conference Proceeding)

TY - GEN

T1 - UnSNAP

T2 - A Mini-App for Exploring the Performance of Deterministic Discrete Ordinates Transport on Unstructured Meshes

AU - Deakin, Tom

AU - McIntosh-Smith, Simon

AU - Lovegrove, Justin

AU - Smedley-Stevenson, Richard

AU - Hagues, Andrew

PY - 2018/11/1

Y1 - 2018/11/1

N2 - Solving the deterministic discrete ordinates neutral particle transport equation is a computationally expensive application. On an unstructured mesh, the discontinuous Galerkin finite element method is used for discretisation of the spatial domain. Additionally, an upwind dependency is applied forming wavefront sweeps across the spatial mesh for each iteration of the solve. We present a new mini-app, UnSNAP, which can be used to investigate the performance of arbitrarily high-order finite element unstructured transport on modern architectures. A new schedule appropriate for such architectures is presented. Finally, we show performance results for the mini-app on CPUs with high numbers of cores.

AB - Solving the deterministic discrete ordinates neutral particle transport equation is a computationally expensive application. On an unstructured mesh, the discontinuous Galerkin finite element method is used for discretisation of the spatial domain. Additionally, an upwind dependency is applied forming wavefront sweeps across the spatial mesh for each iteration of the solve. We present a new mini-app, UnSNAP, which can be used to investigate the performance of arbitrarily high-order finite element unstructured transport on modern architectures. A new schedule appropriate for such architectures is presented. Finally, we show performance results for the mini-app on CPUs with high numbers of cores.

KW - deterministic discrete ordinates transport

KW - finite element

KW - discontinuous Galerkin

KW - sweep

KW - unstructured mesh

U2 - 10.1109/CLUSTER.2018.00076

DO - 10.1109/CLUSTER.2018.00076

M3 - Conference Contribution (Conference Proceeding)

SN - 9781538683200

SP - 598

EP - 606

BT - Workshop of Representative Applications (WRAp) 18 at 2018 IEEE International Conference on Cluster Computing (CLUSTER 2018)

PB - IEEE Computer Society

ER -

Deakin T , McIntosh-Smith S, Lovegrove J, Smedley-Stevenson R, Hagues A. UnSNAP: A Mini-App for Exploring the Performance of Deterministic Discrete Ordinates Transport on Unstructured Meshes. In Workshop of Representative Applications (WRAp) 18 at 2018 IEEE International Conference on Cluster Computing (CLUSTER 2018). IEEE Computer Society. 2018. p. 598-606 Epub 2018 Nov 1. doi: 10.1109/CLUSTER.2018.00076

UnSNAP: A Mini-App for Exploring the Performance of Deterministic Discrete Ordinates Transport on Unstructured Meshes

Abstract

Publication series

Keywords

Access to Document

Handle.net

Fingerprint

Cite this