Loading...
Loading...

Go to the content (press return)

Performance analysis of a mimetic finite difference scheme for acoustic problems: CPU versus GPU computing

Author
Rodriguez, R.; Otero, B.; Frances, J.; Solano, F.; Rojas, O.; Guevara-Jordan, J.
Type of activity
Presentation of work at congresses
Name of edition
VIII Pan-American workshop
Date of publication
2014
Presentation's date
2014-07-21
Book of congress proceedings
VIII Pan-American Workshop: Applied and computational mathematics
First page
1
Last page
1
URL
http://www.csrc.sdsu.edu/panam2014/index.html Open in new window
Abstract
Nowadays, numerical simulations of seismic propagation phenomena demand high performance computing because of the large study domain of interest and the high volume of input/output data. In the case of finite difference methods (or any other volume discretization scheme), processing time might become strikingly large if we additionally use well refined grids in order to capture small physical devices on wave patterns. In this paper, we present several C implementations of a mimetic finite diffe...
Keywords
GPU, acoustic problem, mimetic method
Group of research
VIRTUOS - Virtualisation and Operating Systems

Participants

  • Rodriguez Hernandez, Roberto  (author and speaker )
  • Otero Calviño, Beatriz  (author and speaker )
  • Frances, Jorge  (author and speaker )
  • Solano, Freysimar  (author and speaker )
  • Rojas, Otilio  (author and speaker )
  • Guevara-Jordan, Juan  (author and speaker )