Abstract

International audience; Modern embedded processors have gone through multiple internal optimization to speed-up the average execution time e.g., caches, pipelines, branch prediction. Besides, internal communication mechanisms and shared resources like caches or buses have a significant impact on Worst-Case Execution Times (WCETs). Having an accurate estimate of a WCET is now a challenge. Probabilistic approaches provide a viable alternative to single WCET estimation. They consider WCET as a probabilistic distribution associated to uncertainty or risk.In this paper, we present synthetic benchmarks and associated analysis for several LEON3 configurations on FPGA targets. Benchmarking exposes key parameters to execution time variability allowing for accurate probabilistic modeling of system dynamics. We analyze the impact of architecture-level configurations on average and worst-case behaviors.; Les processeurs embarqués modernes sont passés par optimisation interne multiple pour accélérer l'exécution moyenne par exemple le temps, les caches, les pipelines, la prédiction de branchement. Par ailleurs, mécanismes de communication et de partage des ressources comme des caches ou bus ont un impact significatif sur le temps d'exécution pire des cas (WCETs). Avoir une estimation précise d'un WCET est maintenant en défi. Les approches probabilistes fournies pour des alternatives viables à Singles WCET estimation. Ils considèrent comme un WCET probabiliste la distribution associée à l'incertitude ou le risque du pire cas.Dans cet article, nous présentons des benchmarks synthétiques et associés analyse pour plusieurs configurations LEON3 sur FPGA cibles. Benchmarking expose l'effet de certains paramètres sur la variabilité de l'exécution des taches afin de permettre à la modélisation probabiliste précise de la dynamique du système. Nous analysons l'impact des configurations sur les comportements moyens et pire cas du systeme.

Abstract

Considering that a large part of Sweden is covered by glacial till, which is classified as an unsorted sediment formed by glaciers that can contain fragments of rock known as boulders, driving piles constitutes a substantial economic risk. Piles driven into glacial till may encounter boulders and undergo structural damages leading to premature refusal and to the loss of piles. Even though geotechnical investigations as of today form a solid basis for the design of pile foundations, the unpredictable presence of boulders and their hard resistance to breakage, makes it challenging to penetrate boulders by standard investigation methods. Currently, the only available source of information used by the Swedish construction industry to confirm the existence of boulders is a dynamic penetration test known as soil–rock sounding. Relying on the results from only one testing method may for most projects underestimate the existence of boulders and their potential impact to piles, leading to an unsuitable design of the entire piling system. This paper discusses the benefit in using the input from soil–rock soundings for quantifying the probability of boulder encounters in glacial till based on Poisson point process.