Working on Projects which create high value to our Society

The software we write today is written for an abstract deterministic computer that does not exist. This is not only true when speaking about software written in a high-level programming language, but also even in the case of assembly code. One part that leads to this situation is the tremendous focus on performance, leading to introducing techniques like speculative executions, out-of-order executions, or just merely caches.

This focus on performance has a tremendous implication on the security of the software we write. To write secure software, we need to know and rely on the behavior of our hardware. In this project, we consider hardware with a focus on security first and performance second. We conduct research on a better understanding of the assumption that secure software makes on the hardware and which misjudgments enable devastating attacks. In the end, we want to design hardware that can provide the necessary and sufficient strong security guarantees that secure software needs to function correctly and the software tools required to enable it.