Skip to Main content Skip to Navigation
Theses

Parallélisation et passage à l'échelle durable d'une chaîne de traitement graphique pour l'impression professionnelle

Paul Godard 1, 2, 3
1 CAMUS - Compilation pour les Architectures MUlti-coeurS
Inria Nancy - Grand Est, ICube - Laboratoire des sciences de l'ingénieur, de l'informatique et de l'imagerie
Abstract : The strong and continuous improvements in the professional printing field have led to the ubiquity of printed objects in our daily life. The flexibility introduced by the digital printing process promises to associate extensive customization with mass production. The quick growth of printing usages and technologies, illustrated by wider printer farms and faster digital presses, leads to original challenges for the computer system in charge of driving them. In this thesis, we explore new approaches inspired by the high performance computing field to speedup the graphics processing necessary to digital printing. To achieve this goal, we introduce a distributed system which provides the adequate flexibility and performance by exploiting and optimizing both processing and synchronization techniques. We present our architecture up to the subtle parts of its implementation which allows our solution to meet the specific constraints on generating streams for printing purpose. We perform a complete evaluation of our solution and provide experimental evidence of its great performance and viability.
Complete list of metadata

Cited literature [220 references]  Display  Hide  Download

https://hal.inria.fr/tel-02456793
Contributor : Vincent Loechner <>
Submitted on : Thursday, January 30, 2020 - 11:31:03 AM
Last modification on : Friday, October 23, 2020 - 4:37:18 PM

Identifiers

  • HAL Id : tel-02456793, version 2

Citation

Paul Godard. Parallélisation et passage à l'échelle durable d'une chaîne de traitement graphique pour l'impression professionnelle. Calcul parallèle, distribué et partagé [cs.DC]. Université de strasbourg, 2019. Français. ⟨tel-02456793v2⟩

Share

Metrics

Record views

90

Files downloads

413