Acceleration of TM cylinder EFIE with CUDA

Acceleration of TM cylinder EFIE with CUDA

In this paper, we have shown that exploitation of the GPU's massively parallel architecture can dramatically increase the speed of MoM calculations. While the code can certainly be improved, matrix fill speed-up factors are already commonly found to be between 150X-260X. The conjugate gradient...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Killian, T., Faircloth, D.L., Rao, S.M.
Format: Tagungsbericht
Sprache:eng
Schlagworte:
Acceleration
Central Processing Unit
Computer architecture
Computer graphics
Finite difference methods
Hardware
Optical scattering
Physical optics
Time domain analysis
Yarn
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:In this paper, we have shown that exploitation of the GPU's massively parallel architecture can dramatically increase the speed of MoM calculations. While the code can certainly be improved, matrix fill speed-up factors are already commonly found to be between 150X-260X. The conjugate gradient solver stands to be improved at this writing but still results in performance increases of 1.25X-20X. Continued development of MoM codes with GPU acceleration will undoubtedly bring about a new era of computational electromagnetics.
ISSN:1522-3965
1947-1491
DOI:10.1109/APS.2009.5171648