Trustworthy Computing

Computers you can trust. Whether in a factory, a vehicle, or a medical device, today's applications must work reliably, safely, and securely, even when faults occur or someone attacks.

What we work on

• Reliable computing: detecting and overcoming faults, errors and failures
• Secure computing: closing security gaps in processors and AI accelerators, and designing hardware roots of trust.
• Computation on encrypted data:   Fully Homomorphic Encryption for ML.  

Our goal: trustworthiness that starts in the hardware.

Trustworthiness, designed from the hardware up

A computer is trustworthy when its behaviour can be predicted accurately, that is, when the risks, hazards and threats that could lead to unwanted behaviour are kept small. That takes a high level of security, reliability and safety at the same time. We develop hardware-based solutions for this: trustworthy computing platforms, security by design and reliability in electronics.

In particular, we work on secure and safe processors, efficient cryptographic accelerators, , and verifying the function and security of whole computing platforms. The research field is led by Dr.-Ing. Saleh Mulhem.

Four lines of work

• Trustworthy computing systems: secure RISC-V processors, trustworthy AI accelerators, cryptographic accelerators and hardware roots of trust. More at Secure Computing.
• Design for hardware trust: protecting integrated circuits that are made along a supply chain with many partly untrusted parties, for example against hardware Trojans, cloning and reverse engineering.
• System models for trustworthiness analysis: SystemC models of processors and accelerators on which fault propagation, security level and safety can be studied early in the design.
• Reliability in electronics: assessing and improving reliable accelerators and systems-on-chip. More at Reliable Computing.