Performance Evaluation of Heterogeneous GPU Programming Frameworks for Hemodynamic Simulations

Aristotle Martin; Geng Liu; William Ladd; Seyong Lee; John Gounley; Jeffrey Vetter; Saumil Patel; Silvio Rizzi; Victor Mateevitsi; Joseph Insley; Amanda Randles

doi:10.1145/3624062.3624188

Performance Evaluation of Heterogeneous GPU Programming Frameworks for Hemodynamic Simulations

Aristotle Martin, Geng Liu, William Ladd, Seyong Lee, John Gounley, Jeffrey Vetter, Saumil Patel, Silvio Rizzi, Victor Mateevitsi, Joseph Insley, Amanda Randles

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

5 Scopus citations

Abstract

Preparing for the deployment of large scientific and engineering codes on upcoming exascale systems with GPU-dense nodes is made challenging by the unprecedented diversity of device architectures and heterogeneous programming models. In this work, we evaluate the process of porting a massively parallel, fluid dynamics code written in CUDA to SYCL, HIP, and Kokkos with a range of backends, using a combination of automated tools and manual tuning. We use a proxy application along with a custom performance model to inform the results and identify additional optimization strategies. At scale performance of the programming model implementations are evaluated on pre-production GPU node architectures for Frontier and Aurora, as well as on current NVIDIA device-based systems Summit and Polaris. Real-world workloads representing 3D blood flow calculations in complex vasculature are assessed. Our analysis highlights critical trade-offs between code performance, portability, and development time.

Original language	English
Title of host publication	Proceedings of 2023 SC Workshops of the International Conference on High Performance Computing, Network, Storage, and Analysis, SC Workshops 2023
Publisher	Association for Computing Machinery
Pages	1126-1137
Number of pages	12
ISBN (Electronic)	9798400707858
DOIs	https://doi.org/10.1145/3624062.3624188
State	Published - Nov 12 2023
Event	2023 International Conference on High Performance Computing, Network, Storage, and Analysis, SC Workshops 2023 - Denver, United States Duration: Nov 12 2023 → Nov 17 2023

Publication series

Name	ACM International Conference Proceeding Series

Conference

Conference	2023 International Conference on High Performance Computing, Network, Storage, and Analysis, SC Workshops 2023
Country/Territory	United States
City	Denver
Period	11/12/23 → 11/17/23

Funding

Research reported in this work was supported by the National Institutes of Health under Award Numbers U01CA253511 and T32GM144291, and the ALCF Aurora Early Science Program. The content does not necessarily represent the official views of the NIH. This research used resources from the Argonne Leadership Computing Facility, which is a DOE Office of Science User Facility supported by the DE-AC02-06CH11357 Contract. An award of compute time was provided by the INCITE program. This research used resources from the Oak Ridge Leadership Computing Facility, which is a DOE Office of Science User Facility supported under Contract DE-AC05-00OR22725.

Keywords

Computational fluid dynamics
Performance portability
Proxy applications

Access to Document

10.1145/3624062.3624188

Cite this

Martin, A., Liu, G., Ladd, W., Lee, S., Gounley, J., Vetter, J., Patel, S., Rizzi, S., Mateevitsi, V., Insley, J., & Randles, A. (2023). Performance Evaluation of Heterogeneous GPU Programming Frameworks for Hemodynamic Simulations. In Proceedings of 2023 SC Workshops of the International Conference on High Performance Computing, Network, Storage, and Analysis, SC Workshops 2023 (pp. 1126-1137). (ACM International Conference Proceeding Series). Association for Computing Machinery. https://doi.org/10.1145/3624062.3624188

Martin, Aristotle ; Liu, Geng ; Ladd, William et al. / Performance Evaluation of Heterogeneous GPU Programming Frameworks for Hemodynamic Simulations. Proceedings of 2023 SC Workshops of the International Conference on High Performance Computing, Network, Storage, and Analysis, SC Workshops 2023. Association for Computing Machinery, 2023. pp. 1126-1137 (ACM International Conference Proceeding Series).

@inproceedings{24023e113dcc4b078d1431e9e520a222,

title = "Performance Evaluation of Heterogeneous GPU Programming Frameworks for Hemodynamic Simulations",

abstract = "Preparing for the deployment of large scientific and engineering codes on upcoming exascale systems with GPU-dense nodes is made challenging by the unprecedented diversity of device architectures and heterogeneous programming models. In this work, we evaluate the process of porting a massively parallel, fluid dynamics code written in CUDA to SYCL, HIP, and Kokkos with a range of backends, using a combination of automated tools and manual tuning. We use a proxy application along with a custom performance model to inform the results and identify additional optimization strategies. At scale performance of the programming model implementations are evaluated on pre-production GPU node architectures for Frontier and Aurora, as well as on current NVIDIA device-based systems Summit and Polaris. Real-world workloads representing 3D blood flow calculations in complex vasculature are assessed. Our analysis highlights critical trade-offs between code performance, portability, and development time.",

keywords = "Computational fluid dynamics, Performance portability, Proxy applications",

author = "Aristotle Martin and Geng Liu and William Ladd and Seyong Lee and John Gounley and Jeffrey Vetter and Saumil Patel and Silvio Rizzi and Victor Mateevitsi and Joseph Insley and Amanda Randles",

note = "Publisher Copyright: {\textcopyright} 2023 ACM.; 2023 International Conference on High Performance Computing, Network, Storage, and Analysis, SC Workshops 2023 ; Conference date: 12-11-2023 Through 17-11-2023",

year = "2023",

month = nov,

day = "12",

doi = "10.1145/3624062.3624188",

language = "English",

series = "ACM International Conference Proceeding Series",

publisher = "Association for Computing Machinery",

pages = "1126--1137",

booktitle = "Proceedings of 2023 SC Workshops of the International Conference on High Performance Computing, Network, Storage, and Analysis, SC Workshops 2023",

}

Martin, A, Liu, G, Ladd, W, Lee, S , Gounley, J , Vetter, J, Patel, S, Rizzi, S, Mateevitsi, V, Insley, J & Randles, A 2023, Performance Evaluation of Heterogeneous GPU Programming Frameworks for Hemodynamic Simulations. in Proceedings of 2023 SC Workshops of the International Conference on High Performance Computing, Network, Storage, and Analysis, SC Workshops 2023. ACM International Conference Proceeding Series, Association for Computing Machinery, pp. 1126-1137, 2023 International Conference on High Performance Computing, Network, Storage, and Analysis, SC Workshops 2023, Denver, United States, 11/12/23. https://doi.org/10.1145/3624062.3624188

Performance Evaluation of Heterogeneous GPU Programming Frameworks for Hemodynamic Simulations. / Martin, Aristotle; Liu, Geng; Ladd, William et al.
Proceedings of 2023 SC Workshops of the International Conference on High Performance Computing, Network, Storage, and Analysis, SC Workshops 2023. Association for Computing Machinery, 2023. p. 1126-1137 (ACM International Conference Proceeding Series).

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

TY - GEN

T1 - Performance Evaluation of Heterogeneous GPU Programming Frameworks for Hemodynamic Simulations

AU - Martin, Aristotle

AU - Liu, Geng

AU - Ladd, William

AU - Lee, Seyong

AU - Gounley, John

AU - Vetter, Jeffrey

AU - Patel, Saumil

AU - Rizzi, Silvio

AU - Mateevitsi, Victor

AU - Insley, Joseph

AU - Randles, Amanda

PY - 2023/11/12

Y1 - 2023/11/12

N2 - Preparing for the deployment of large scientific and engineering codes on upcoming exascale systems with GPU-dense nodes is made challenging by the unprecedented diversity of device architectures and heterogeneous programming models. In this work, we evaluate the process of porting a massively parallel, fluid dynamics code written in CUDA to SYCL, HIP, and Kokkos with a range of backends, using a combination of automated tools and manual tuning. We use a proxy application along with a custom performance model to inform the results and identify additional optimization strategies. At scale performance of the programming model implementations are evaluated on pre-production GPU node architectures for Frontier and Aurora, as well as on current NVIDIA device-based systems Summit and Polaris. Real-world workloads representing 3D blood flow calculations in complex vasculature are assessed. Our analysis highlights critical trade-offs between code performance, portability, and development time.

AB - Preparing for the deployment of large scientific and engineering codes on upcoming exascale systems with GPU-dense nodes is made challenging by the unprecedented diversity of device architectures and heterogeneous programming models. In this work, we evaluate the process of porting a massively parallel, fluid dynamics code written in CUDA to SYCL, HIP, and Kokkos with a range of backends, using a combination of automated tools and manual tuning. We use a proxy application along with a custom performance model to inform the results and identify additional optimization strategies. At scale performance of the programming model implementations are evaluated on pre-production GPU node architectures for Frontier and Aurora, as well as on current NVIDIA device-based systems Summit and Polaris. Real-world workloads representing 3D blood flow calculations in complex vasculature are assessed. Our analysis highlights critical trade-offs between code performance, portability, and development time.

KW - Computational fluid dynamics

KW - Performance portability

KW - Proxy applications

UR - https://www.scopus.com/pages/publications/85178118548

U2 - 10.1145/3624062.3624188

DO - 10.1145/3624062.3624188

M3 - Conference contribution

AN - SCOPUS:85178118548

T3 - ACM International Conference Proceeding Series

SP - 1126

EP - 1137

BT - Proceedings of 2023 SC Workshops of the International Conference on High Performance Computing, Network, Storage, and Analysis, SC Workshops 2023

PB - Association for Computing Machinery

T2 - 2023 International Conference on High Performance Computing, Network, Storage, and Analysis, SC Workshops 2023

Y2 - 12 November 2023 through 17 November 2023

ER -

Martin A, Liu G, Ladd W, Lee S , Gounley J , Vetter J et al. Performance Evaluation of Heterogeneous GPU Programming Frameworks for Hemodynamic Simulations. In Proceedings of 2023 SC Workshops of the International Conference on High Performance Computing, Network, Storage, and Analysis, SC Workshops 2023. Association for Computing Machinery. 2023. p. 1126-1137. (ACM International Conference Proceeding Series). doi: 10.1145/3624062.3624188

Performance Evaluation of Heterogeneous GPU Programming Frameworks for Hemodynamic Simulations

Abstract

Publication series

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this