Artikel in Tagungsband INPROC-2016-41

Schnitzer, Stephan; Gansel, Simon; Dürr, Frank; Rothermel, Kurt: Real-time scheduling for 3D GPU rendering.
In: 11th IEEE International Symposium on Industrial Embedded Systems (SIES).
Universität Stuttgart, Fakultät Informatik, Elektrotechnik und Informationstechnik.
S. 1-10, englisch.
IEEE, Mai 2016.
DOI: 10.1109/SIES.2016.7509411.
Artikel in Tagungsband (Konferenz-Beitrag).
CR-Klassif.H.5.2 (Information Interfaces and Presentation User Interfaces)
I.3.m (Computer Graphics Miscellaneous)
Keywordsautomobiles; graphical user interfaces; graphics processing units; rendering (computer graphics); 3D GPU rendering; GPU sharing; real-time GPU scheduling framework; real-time scheduling; automotive HMI; embedded systems

3D graphical functions in cars enjoy growing popularity. For instance, analog instruments of the instrument cluster are replaced by digital 3D displays as shown by Mercedes-Benz in the F125 prototype car. The trend to use 3D applications expands into two directions: towards more safety-relevant applications such as the speedometer and towards third-party applications, e.g., from an app store. In order to save cost, energy, and installation space, all these applications should share a single GPU. GPU sharing brings up the problem of providing real-time guarantees for rendering content of time-sensitive applications like the speedometer. To solve this problem, we present a real-time GPU scheduling framework which provides strong guarantees for critical applications while still giving as much GPU resources to less important applications as possible, thus ensuring a high GPU utilization. Since current GPUs are not preemptible, we use the estimated execution time of each GPU rendering job to make the scheduling decisions. Our evaluations show that our scheduler guarantees given real-time constraints, while achieving a high GPU utilization of 97%. Moreover, scheduling is performed highly efficient in real-time with less than 10 us latency.

Volltext und
andere Links
PDF (1292181 Bytes)
The original publication is available at IEEE Xplore
Copyright© 2016 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.
Abteilung(en)Universität Stuttgart, Institut für Parallele und Verteilte Systeme, Verteilte Systeme
Eingabedatum26. November 2016
   Publ. Abteilung   Publ. Institut   Publ. Informatik