Soutenance de thèse de Julien Plante le jeudi 30 novembre 2023

Titre de la thèse

"Enabling technologies for real-time acquisition and processing of large volumes of data and their applications to giant astronomical telescopes and radar systems"

Résumé

Systems of our world are following a constant growth in terms of precision, speed and electrical consumption. This growth involves the production of always bigger volumes of data, which become difficult to process using today’s standard technologies. In the context of this PhD, we looked into emerging technologies such as GPU computing, userland networking, DPDK and GPUDirect, which are able to answer this need, and applied them to multiple real-life projects (wavefront acquisition for adaptive optics, real-time fast radio burst detection, radar signal processing) with success. These results are very encouraging towards overcoming the bottleneck of current standard technologies, and are promising regarding a massive use in many domains (astronomy, radar, but also autonomous vehicules, finance, etc).