GPGPU and MIC in Accelerated Cluster for Remote Sensed Image Processing Software
CNES
Conference on Big Data from Space (BiDS), 2015
@article{melet2015gpgpu,
title={GPGPU and MIC in Accelerated Cluster for Remote Sensed Image Processing Software},
author={MELET, Olivier and Architect, Ground Segment},
year={2015}
}
Processing of Earth observation remotely sensed images requires more and more powerful computing facilities. Since a few years, GPGPU (General Purpose processing on Graphics Processing Units) technology has been used to perform massively parallel calculations. The French Space Agency (CNES) has then made a portage of some IAS to assess their performance using this type of technology. Based on the very encouraging results and in order to improve operations during the in orbit commissioning phases of Earth observation satellites, CNES is developing, in partnership with Thales Services, an Expertise and Image Calibration Center based on a computer cluster accelerated by such technologies. Starting from an overall description of the image calibration objectives and GPGPU applications, the paper focuses on the study leaded by CNES in partnership with the Antwerp University and Thales Services to identify the most appropriate technologies to define a full computing cluster architecture.
February 3, 2015 by hgpu