Together with my colleages at the Barkhausen Institut, I am continuing the research on the new system architecture I introduced in my dissertation. My disseration layed the foundation for a secure composition of arbitrary compute units, but focus on heterogeneity. Now that the foundation is available, we use the system architecture to investigate how the security of systems can be improved. The basic idea is to physically isolate compute units from each other based on the new hardware component introduced by the system architecture. This allows us to go further than existing approaches, which typically assume that the hardware can be trusted.

In my dissertation, I introduced a new system architecture for heterogeneous compute units. The goal was to support compute units ranging from complex general purpose cores to fixed-function accelerators as first-class citizens by providing spatial isolation, direct communication, and access to OS services for all compute units. The key idea of this approach is to introduce a new hardware component next to each compute unit as a uniform interface for all compute units. Besides this hardware component, I designed a new operating system that takes advantage of the uniform interface to manage very heterogeneoussystems easily and efficiently.

Teaching of software engineering classes, creation of a prime example for students with Apache Wicket, and modelling of dynamic systems in Alloy.