Figure 1
Grid partitioning of the atmopshere in the global model. Each GPU thread is assigned the calculation for one cell, stacked in arrays of columns.
Figure 2
Tasks offload execution GPU.
MEDINA: MECCA Development in Accelerators – KPP Fortran to CUDA source-to-source Pre-processor
DOI: https://doi.org/10.5334/jors.158 | Journal eISSN: 2049-9647
Language: English
Submitted on: Dec 9, 2016
|Accepted on: Apr 20, 2017
|Published on: Apr 28, 2017
Published by: Ubiquity Press
In partnership with: Paradigm Publishing Services
Publication frequency: 1 issue per year
Keywords:
© 2017 Michail Alvanos, Theodoros Christoudias, published by Ubiquity Press
This work is licensed under the Creative Commons Attribution 4.0 License.