RT Conference Proceedings
T1 Weather Projections and Dynamical Downscaling for the Republic of Panama: Evaluation of Implementation Methods via GPGPU Acceleration
A1 Pinzón, Reinhardt
A1 Müller, Michel
A1 Nakaegawa, Toshiyuki
A1 Sánchez-Galán, Javier
A1 Ujaldon-Martínez, Manuel
A1 Fábrega, José
K1 Tiempo (Meteorología) - Previsión
AB Climate change could have a critical impact on the Republic of Panama where a major segment of the economy is dependenton the operation of the Panama Canal. New capabilities to do targeted research around climate change impacts on Panama istherefore being established. This includes anew GPU-cluster infrastructure called Iberogun, based around 2 DGX1 servers(each running 16 NVIDIA Tesla P100 GPUs). This infrastructure will be used to evaluate potential climate models and modelsof extreme weather events. In this review we therefore present an evaluation of the GPGPU (General Purpose GraphicProcessing Unit, here abbreviated GPU) implementation methods for the study of weather projections and dynamicaldownscaling in the Republic of Panama. Different methods are discussed, including: domain-specific languages (DSLs),directive-based porting methods, granularity optimization methods, and memory layout transforming methods. One of theseapproaches that has yielded interesting previous results is further discussed, a directive-based code transformation methodcalled ‘Hybrid Fortran’ that permits a high-performance GPU port for arranged lattice Fortran codes. Finally, we suggest amethod akin to previous investigations related to climate change done for the Republic of Panama, but with acceleration viaGPU capabilities.
YR 2020
FD 2020-09
LK https://hdl.handle.net/10630/30308
UL https://hdl.handle.net/10630/30308
LA eng
NO R. Pinzón, M. Muller, R. Nakaegawa, J. Sánchez-Galán, J. Fábrega, M. Ujaldón. "Weather Projections and Dynamical Downscaling for the Republic of Panama: Evaluation of Implementation Methods via GPGPU Acceleration". Proceedings 22nd Congress of Intl. Association for Hydro Environment Engineering and Research (IAHR) and Asia Pacific Division (APD). Sapporo (Japón), 14 al 17 de Septiembre de 2020.
NO We acknowledge a scientific fund from Sistema Nacional de Investigación de Panamá (SNI) and Projects: FID-2016-275 and EIE-2018-16 of Convocatorias públicas of Secretaria Nacional de Ciencia y Tecnología e Innovación(SENACYT). We acknowledge funds and support from JSPS Grant-in-Aid for Specially Promoted Research16H06291. We acknowledge Theme C of the TOUGOU program granted by the Japanese Ministry ofEducation, Culture, Sports, Science and Technology. The authors thank the Universidad Tecnológica de Panamátheir extensive support, and for the use of their CIHH-group HPC-Cluster-Iberogun. Also acknowledge to NVIDIACorporation with the donation of the Titan Xp GPU used for this research.
DS RIUMA. Repositorio Institucional de la Universidad de Málaga
RD 20 ene 2026