Alya towards Exascale: Optimal OpenACC Performance of the Navier-Stokes Finite Element Assembly on GPUs

Owen H, Ernst D, Gruber T, Lemkuhl O, Houzeaux G, Gasparino L, Wellein G (2024)

Publication Language: English

Publication Status: Published

Publication Type: Conference contribution, Conference Contribution

Publication year: 2024

Publisher: Institute of Electrical and Electronics Engineers Inc.

Pages Range: 408-416

Conference Proceedings Title: 2024 IEEE International Parallel and Distributed Processing Symposium (IPDPS)

Event location: San Francisco US

ISBN: 9798350337662

DOI: 10.1109/IPDPS57955.2024.00043

Abstract

This paper addresses the challenge of providing portable and highly efficient code structures for CPU and GPU architectures. We choose the assembly of the right-hand term in the incompressible flow module of the High-Performance Computational Mechanics code Alya, which is one of the two CFD codes in the Unified European Benchmark Suite. Starting from an efficient CPU-code and a related OpenACC-port for GPUs we successively investigate performance potentials arising from code specialization, algorithmic restructuring and low-level optimizations.We demonstrate that only the combination of these different dimensions of runtime optimization unveils the full performance potential on the GPU and CPU. Roofline-based performance modelling is applied in this process and we demonstrate the need to investigate new optimization strategies if a classical roofline limit such as memory bandwidth utilization is achieved, rather than stopping the process. The final unified OpenACC-based implementation boosts performance by more than 50x on an NVIDIA A100 GPU (achieving approximately 2.5 TF/s FP64) and a further factor of 5x for an Intel Icelake based CPU-node (achieving approximately 1.0 TF/s FP64).The insights gained in our manual approach lays ground implementing unified but still highly efficient code structures for related kernels in Alya and other applications. These can be realized by manual coding or automatic code generation frameworks.

Authors with CRIS profile

Dominik Ernst Erlangen National High Performance Computing Center (NHR@FAU) Thomas Gruber Erlangen National High Performance Computing Center (NHR@FAU) Gerhard Wellein Professur für Höchstleistungsrechnen

Additional Organisation(s)

Erlangen National High Performance Computing Center (NHR@FAU)

Related research project(s)

FOSTERING THE EUROPEAN ENERGY TRANSITION WITH EXASCALE (EoCoE-III) FOSTERING THE EUROPEAN ENERGY TRANSITION WITH EXASCALE Jan. 1, 2024 - Dec. 31, 2026

Involved external institutions

Barcelona Supercomputing Center / Centro Nacional de Supercomputación ES Spain (ES)

How to cite

APA:

Owen, H., Ernst, D., Gruber, T., Lemkuhl, O., Houzeaux, G., Gasparino, L., & Wellein, G. (2024). Alya towards Exascale: Optimal OpenACC Performance of the Navier-Stokes Finite Element Assembly on GPUs. In 2024 IEEE International Parallel and Distributed Processing Symposium (IPDPS) (pp. 408-416). San Francisco, US: Institute of Electrical and Electronics Engineers Inc..

MLA:

Owen, Herbert, et al. "Alya towards Exascale: Optimal OpenACC Performance of the Navier-Stokes Finite Element Assembly on GPUs." Proceedings of the 38th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2024, San Francisco Institute of Electrical and Electronics Engineers Inc., 2024. 408-416.

BibTeX: Download