No abstract

Parallel computing has been aniche for scientific research in academia for decades. However, as common industrial applications become more and more performance demanding and raising the clock frequencyofconventional single-core systems is hardly an option due to reaching technological limitations, efficient use of (embedded) multi-core CPUs and many-core platforms has become imperative.3 Ds urface analysis of objects using the white light interferometry presents one of such challenging applications. The goal in this article is to get an impression which speed-up for an established and parallelized white light interferometry preprocessing algorithm, called Contrast Method,i spossible on an embedded system that works without anyoperating system. Therefore, we decided to use avirtual environment that is able to simulate embedded multi-core as well as many-core systems and that enables running real application code on the designed system. The results show, that asignificant speed-up is possible when using amany-core platform, instead of adesign that only implements one single core, if the algorithm is parallelized for getting full advantage of the manycore design. Furthermore, an acceptable absolute run time is achievable.

This paper presents the new Framework for Real-time capable Embedded system and ArChitecture SIMulation (FREACSIM), a highly configurable full-system simulation environment enabling and easing the modeling, simulation and verification of Network on Chip architectures for hard real-time systems. The framework is mostly geared towards software developers, supporting them in the simulation of NoCs at an instruction accurate level and offers a broad variety of real-world hardware components as part of the integrated virtualization toolbox. FREACSIM provides a software-based routing strategy between nodes, with a single node consisting of a processor and required peripherals for enabling real-time capability. This allows a flexible and independent comparison of currently implemented hardware strategies, as well as an easy adaption to better suit new hardware needs. The softwarebased routing, as well as distributed applications that can be implemented for the NoC hardware design, are able to use the real-time operating system eCos, which is part of our framework. As a result, real-time capable software can be implemented for, and tested on, complex NoC systems. We demonstrate the flexibility and the benefits of our framework with a set of applications (use cases), which cover typical heavy and light load distributions between communication and computation.

No abstract

No abstract