In the PORTUNUS project, we are researching cryptoagile APIs for hardware-implemented post-quantum cryptographyas part of the ATHENE research area Cryptography .

Progress in the development of cryptographic methods that are resistant to quantum computers is progressing rapidly. The number of cryptographic algorithms has risen continuously in recent years. This increase is due to various events, including in particular the process initiated by NIST to identify post-quantum algorithms. In addition, there is a large number of variants of cryptographic algorithms and cryptographic libraries.

For efficient, practical use, a triad between cryptographers who develop PQC algorithms, experts for platform integration who implement these algorithms on the hardware and software developers who want to use them must be harmonised. The partners at TU Darmstadt and Darmstadt University of Applied Sciences are working on this topic in the PORTUNUS project.

In the context of this topic, TU Darmstadt is focussing on the theoretical question of how emerging PQC algorithms can be accelerated more effectively. This essentially involves the development of multi-level intermediate representation MLIR compiler optimisations for CPU, GPU, ASICs and FPGAs.

Darmstadt University of Applied Sciences is developing a platform for integrating these compiler-optimised algorithms. For this purpose, an API is being developed that enables flexible and user-friendly use of these algorithms under the specified cryptographic requirements.

In the context of cryptoagility, the API offers a set of cryptographic libraries (such as OQS and OpenSSL) from which any cryptographic algorithms can be selected depending on the application.

The API is subdivided into several modules, each of which performs specific tasks in the encryption process. These include key storage, the management of information on the algorithms and the selection of algorithms by the API.

The aim is to abstract the complexity of cryptography in order to make it easier for developers to handle cryptographic components while minimising implementation errors and the associated security risks.