Wavefront phase recovery using graphic processing units (GPUs)

Fernando L. Rosa; Jose Gil Marichal-Hernandez; Jose Manuel Rodriguez-Ramos

doi:10.1117/12.564777

11 November 2004 Wavefront phase recovery using graphic processing units (GPUs)

Fernando L. Rosa, Jose Gil Marichal-Hernandez, Jose Manuel Rodriguez-Ramos

Proceedings Volume 5572, Optics in Atmospheric Propagation and Adaptive Systems VII; (2004) https://doi.org/10.1117/12.564777
Event: Remote Sensing, 2004, Maspalomas, Canary Islands, Spain

Abstract

We have developed a Shack-Hartmann sensor simulation, moving the complex amplitude of the electromagnetic field using Fast Fourier Transforms. The Shack-Hartmann sensor takes as input the atmospheric wavefront frames generated by the Roddier algorithm, and provides, as output, the subpupil images. The centroids and the wavefront phase maps are computed combining GPU and CPU. The algorithms used on the GPU are written using nVidia language C for Graphics (Cg) and run on a CineFx graphical engine. Such a graphical engine provides a computational power several times greater than usual CPU-FPU combination, with a reduced cost. Any algorithm implemented on these engines must be previously adapted from their original form to fit the pipeline capabilities. To achieve an optimal performance, we compare the results with the same algorithm implemented on GPU and CPU. We present here, for the first time, preliminary results on wavefront phase recovery using GPU. We have chose a zonal algorithm that fits better on the stream paradigm of the GPU's. The result shows a 10x speedup in the GPU centroid algorithm implementation and a 2x speedup in the phase recovery one compared with the same on CPU.

Citation Download Citation

Fernando L. Rosa, Jose Gil Marichal-Hernandez, and Jose Manuel Rodriguez-Ramos "Wavefront phase recovery using graphic processing units (GPUs)", Proc. SPIE 5572, Optics in Atmospheric Propagation and Adaptive Systems VII, (11 November 2004); https://doi.org/10.1117/12.564777

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available