Bachelor and Master Theses

Title: Dynamic Load Generator: Synthesising dynamic hardware load characteristics
Subject: Computer Science
Level: Advanced
Description: Performance penalties that arise from updating the operating system for large complex
systems are diffcult to predict early in the development cycle. Industrial practice is to
measure the system performance at the end of the development cycle. This is not very
effcient in the development of large systems since the cycles can often span over several
months or even years. One way to avoid this problem is to create a model which mimics
hardware characteristics of the production system. These characteristics can be measured
with different metrics, for example cache hit rate, branch hit/miss ratio or interrupt rate.
The model can then be applied to a test node running the updated version of the operating
system and any resulting performance penalties can be detected.
Current research has focused on solving this problem by using static models. The problem
with static models is that they do not take into account how the characteristics of the
system changes over time. An improvement is to create a dynamic model, which is the aim
of this thesis. The diffculty lies in automatically creating such a dynamic model, in reasonable
time, that mimics the production systems as closely as possible. There will probably
be some trade-off between synthesis quality and the time to synthesise, which needs to be
investigated. In "Towards feedback-based generation of hardware characteristics" [1], the
automatisation is solved by using a PID controller. One solution would be to extend this
model such that is becomes dynamic, while at the same time considering synthesis quality
in respect to the time to synthesise.
Company: Ericsson, kontaktperson: Marcus Jägemar
Student: Stefan Karlsson
Student: Erik Hansson
IDT supervisor: Moris Behnam, 021-107094
Examinator: Mikael Sjödin
Mikael Sjödin, +46 70 288 2829

Rapport och bilagor


Senaste uppdatering



2015-05-22, 12:34

  • Mälardalen University |
  • Box 883 |
  • 721 23 Västerås/Eskilstuna |
  • 021-101300, 016-153600 |
  • webmaster |
  • Latest update: 2018.05.24