Congratulations on passing your doctoral examination!

Embedded systems, such as those found in vehicles, must meet certain requirements. Therefore, appropriate simulation environments and tools are needed in the design phase to incorporate these constraints into the search for an optimal system architecture. Distributing the workload across multiple (heterogeneous) computing platforms is a well-known method to enhance performance and energy efficiency. 

In this context, the thesis introduces novel concepts at different design levels, optimizing the hardware architecture with respect to various metrics. The main focus is on inference partitioning at the system level, with the presented framework CNNParted optimizing the distribution of neural networks across available hardware resources. 

The conducted case studies ultimately showed that the proposed resource-based inference partitioning is advantageous compared to current state-of-the-art methods.

Fabian Kreß' research results contribute significantly to the further development of AI-based applications and provide valuable insights into the optimization of system architectures for future technologies. We congratulate him on this significant achievement and wish him all the best for his future scientific and professional endeavors.