A Design-Time/Run-Time Application Map** Methodology for Predictable Execution Time in MPSoCs
Authors:
Andreas Weichslgartner,
Stefan Wildermann,
Deepak Gangadharan,
Michael Glaß,
Jürgen Teich
Abstract:
Executing multiple applications on a single MPSoC brings the major challenge of satisfying multiple quality requirements regarding real-time, energy, etc. Hybrid application map** denotes the combination of design-time analysis with run-time application map**. In this article, we present such a methodology, which comprises a design space exploration coupled with a formal performance analysis.…
▽ More
Executing multiple applications on a single MPSoC brings the major challenge of satisfying multiple quality requirements regarding real-time, energy, etc. Hybrid application map** denotes the combination of design-time analysis with run-time application map**. In this article, we present such a methodology, which comprises a design space exploration coupled with a formal performance analysis. This results in several resource reservation configurations, optimized for multiple objectives, with verified real-time guarantees for each individual application. The Pareto-optimal configurations are handed over to run-time management which searches for a suitable map** according to this information. To provide any real-time guarantees, the performance analysis needs to be composable and the influence of the applications on each other has to be bounded. We achieve this either by spatial or a novel temporal isolation for tasks and by exploiting composable NoCs. With the proposed temporal isolation, tasks of different applications can be mapped to the same resource while with spatial isolation, one computing resource can be exclusively used by only one application. The experiments reveal that the success rate in finding feasible application map**s can be increased by the proposed temporal isolation by up to 30% and energy consumption can be reduced compared to spatial isolation.
△ Less
Submitted 16 November, 2017;
originally announced November 2017.
