Towards a GPU-Parallelization of the neXtSIM-DG Dynamical Core
Fuente:
arXiv
Salvato in:
| Autori principali: | , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866907814650183680 |
|---|---|
| author | Jendersie, Robert Lessig, Christian Richter, Thomas |
| author_facet | Jendersie, Robert Lessig, Christian Richter, Thomas |
| contents | The cryosphere plays a significant role in Earth's climate system. Therefore, an accurate simulation of sea ice is of great importance to improve climate projections. To enable higher resolution simulations, graphics processing units (GPUs) have become increasingly attractive as they offer higher floating point peak performance and better energy efficiency compared to CPUs. However, making use of this theoretical peak performance, which is based on massive data parallelism, usually requires more care and effort in the implementation. In recent years, a number of frameworks have become available that promise to simplify general purpose GPU programming. In this work, we compare multiple such frameworks, including CUDA, SYCL, Kokkos and PyTorch, for the parallelization of \nextsim, a finite-element based dynamical core for sea ice. We evaluate the different approaches according to their usability and performance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2402_00466 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Towards a GPU-Parallelization of the neXtSIM-DG Dynamical Core Jendersie, Robert Lessig, Christian Richter, Thomas Distributed, Parallel, and Cluster Computing Numerical Analysis The cryosphere plays a significant role in Earth's climate system. Therefore, an accurate simulation of sea ice is of great importance to improve climate projections. To enable higher resolution simulations, graphics processing units (GPUs) have become increasingly attractive as they offer higher floating point peak performance and better energy efficiency compared to CPUs. However, making use of this theoretical peak performance, which is based on massive data parallelism, usually requires more care and effort in the implementation. In recent years, a number of frameworks have become available that promise to simplify general purpose GPU programming. In this work, we compare multiple such frameworks, including CUDA, SYCL, Kokkos and PyTorch, for the parallelization of \nextsim, a finite-element based dynamical core for sea ice. We evaluate the different approaches according to their usability and performance. |
| title | Towards a GPU-Parallelization of the neXtSIM-DG Dynamical Core |
| topic | Distributed, Parallel, and Cluster Computing Numerical Analysis |
| url | https://arxiv.org/abs/2402.00466 |