Scholarly record
ASSESSMENT OF GPU COMPUTATIONAL ENHANCEMENT TO WAVE FIELD MODELING
Abstract
In this paper, we present a performance assessment of the implementation of the algorithm of wave sources location using wave field modelling on a Graphics Processing Units (GPUs). The numerical algorithm of wave source location based on the acoustic wave equation was implemented using finite difference method. To achieve better wave source location, the algorithm was implemented using staggered grid meshes. The parallelization was carried out using GPU approach, established in NVIDIAпїЅs Compute Unified Development Architecture (CUDA). Several tests have been made to establish efficiency of the proposed parallelization method. The obtained performance measurement was carried out to determine both: the scalability of a single modelling step and the entire application performance. Tests were performed not only for the different size of the computational model but also for different GPUs cards. This type of testing allowed for reliable assessment of computational enhancement of the GPU implementation of wave field modelling algorithm.
Publication Impact Profile
Publication details
References6
Krauzowicz Ł., Szostek K., Dwornik M, Numerical calculations for geophysics inversion problem using apache hadoop technology. Communications in Computer and Information Science, 291, pp. 440-447, DOI: 10.1007/978-3-642-31217-5_46
Danek T., Seismic Wave Field Modeling with Graphics Processing Units, Lecture Notes in Computer Science vol: 5545, pp. 435-442, 2009
Franczyk A., Gwiżdż D., Leśniak A., Time reversal seismic source imaging using peak average power ratio (PAPR) parameter, Acta Geophys, vol.65/issue 2, pp 299-308, 2017, DOI: 10.1007/s11600-017-0022-0
Videle J.E., Finite-difference traveltime calculation, Seimol. Soc. Am, vol.78/issue 6, pp 2062-2076, 1988, DOI: 10.1190/1.1889870.
Levander A., Fourth-order finite-difference P-SV seismograms, Geophysics, vol.53/issue 11, pp 1425-1436, 1988, DOI: 10.1190/1.1442422.
https://github.com/google/benchmark
View or Download full articleAccess options
SWS access login
Login as SWS Scientific CommitteeLogin as SWS Scientific PartnerLogin as SWS AuthorAuthors and approved SWS contributors will read and export their own linked papers after identity matching by SWS profile, email and SGEM GlobalID.
For librarian assistance: [email protected]
Purchase Instant Access
- Article can be downloaded after successful payment.
- Article may be used according to SWS library access terms.
- Article cannot be redistributed.
