english  | Home | Impressum | Datenschutz | KIT

Investigating Adaptive Networks-on-Chip for Mixed-Criticality Systems

Investigating Adaptive Networks-on-Chip for Mixed-Criticality Systems
Typ:Bachelor-/ Masterarbeit
Datum:offen (zu vergeben)
Betreuer:

M. Sc. Nidhi Anantharajaiah

Investigating Adaptive Networks-on-Chip for Mixed-Criticality Systems

Background

A growing trend in designing embedded systems is that multiple applications with different requirements and criticality share the same hardware platform to improve resource utilization and cost. These platforms are moving towards multi/many core architectures and use Networks-on-Chip (NoC) for communication. With the increasing number of cores on such platforms, the complexity of the interconnect is increasing and it will become a limiting factor for performance.  Efficiency and performance can be improved if the NoC is able to adapt to  different applications while still meeting the requirements. Designing a flexible NoC architecture which can be managed at runtime is a key challenge.

 

 

Tasks

An existing NoC shall be made adaptable by changing the network infrastructure at varying levels of granularity. Some possible examples of such adaptations are: designing flexible topologies, routers and algorithms. The NoC will be designed using Hardware Description Languages, implemented and tested on a multi-FPGA system. The exact task will be decided after consultation with the supervisor.

 

 

Requirements

  • Knowledge of hardware description languages like VHDL or Verilog.
  • Experience with hardware design and with using design tools for Xilinx FPGAs is an advantage.
  • Motivation and interest in solving technical problems independently.