Disclaimer: Dieser Thread wurde aus dem alten Forum importiert. Daher werden eventuell nicht alle Formatierungen richtig angezeigt. Der ursprüngliche Thread beginnt im zweiten Post dieses Threads.
Dolby Nuremberg _Internship – Modeling computational complexity of DSP algorithms
Dear All,
For our client Dolby we are looking for position:
Internship – Modeling computational complexity of DSP algorithms
In order to enable new technologies within next generation devices, such as Digital TV’s and Set Top Boxes, it is essential that the embedded processors that are built into such devices deliver the computational performance that allows these new algorithms to run in real-time.
Because the landscape of embedded processors (IC’s) is quite wide it is not straightforward to understand the exact performance of an audio signal processing algorithm on all the different IC’s.
In this project you’ll be tasked to develop a model that enables engineers within Dolby to predict how the computational complexity of a certain signal processing algorithm (that is characterized a set of parameters) scales on embedded platforms about which certain characteristics are known.
Responsibilities:
•Understand the principle architectures of relevant embedded processors
•Experiment with various algorithms, processors and associated tool-chains in order to develop an empirical or mathematical model that can be used to predict how computational complexity scales with algorithmic complexity.
•Conclude the research investigations by writing a manual on how to use the developed model.
Qualifications:
•Being a student in Electrical Engineering or Computer Science
•Good digital signal processing (DSP) knowledge.
•An understanding of embedded processor architectures
•Practical experiences with performance analysis on embedded platforms (like TI or ARM) is a plus
•Good communications skills in German and/or English.
Please do not hesitate to send cv on: joanna.luczynska@futurestep.com