INVESTIGADORES
STORTI Mario Alberto
artículos
Título:
GPGPU implementation of the BFECC algorithm for pure advection equations
Autor/es:
COSTARELLI, SANTIAGO; STORTI MARIO; PAZ RODRIGO; DALCÍN LISANDRO; IDELSOHN, SERGIO
Revista:
CLUSTER COMPUTING-THE JOURNAL OF NETWORKS SOFTWARE TOOLS AND APPLICATIONS
Editorial:
SPRINGER
Referencias:
Lugar: Berlin; Año: 2014 vol. 17 p. 243 - 254
ISSN:
1386-7857
Resumen:
In the present work an implementation of the Back and ForthError Compensation and Correction (BFECC) algorithm specially suitedfor running on General- Purpose Graphics Processing Units (GPGPUs)through Nvidia?s Compute Unified Device Architecture (CUDA) isanalyzed in order to solve transient pure advection equa- tions. Theobjective is to compare it to a previous ex- plicit version used in aNavier-Stokes solver fully written in CUDA. It turns out that BFECCcould be implemented with unconditional stable stability usingSemi-Lagrangian time integration allowing larger time steps thanEulerian ones.
