Performance Evaluation of Heterogeneous GPU Programming Frameworks for Hemodynamic Simulations
Biomedical Engineering, Duke University, Durham, NC, USA
Proceedings of the SC ’23 Workshops of The International Conference on High Performance Computing, Network, Storage, and Analysis (SC-W’23), 2023
@inproceedings{martin2023performance,
title={Performance Evaluation of Heterogeneous GPU Programming Frameworks for Hemodynamic Simulations},
author={Martin, Aristotle and Liu, Geng and Ladd, William and Lee, Seyong and Gounley, John and Vetter, Jeffrey and Patel, Saumil and Rizzi, Silvio and Mateevitsi, Victor and Insley, Joseph and others},
booktitle={Proceedings of the SC’23 Workshops of The International Conference on High Performance Computing, Network, Storage, and Analysis},
pages={1126–1137},
year={2023}
}
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.
December 31, 2023 by hgpu
