In the era of booming digitalization, our reliance on energy-hungry digital systems continues to surge. This voracious development, coupled with the threat of global warming, necessitates innovative solutions. Approximate Computing (AxC) is a design technique offering an attractive trade-off between computation errors and substantial energy reduction. This PhD proposal aims to push AxC to new frontiers by resorting to advanced semiconductor technologies, including resistive memories, for conceiving probabilistic approximate circuits for cryptographic applications.
The PhD project will include hands-on experimentation with integrated circuits, fabricated at CEA, to characterize and model their probabilistic behaviour. After that, the candidate will propose new AxC techniques and will realize a chip for experimental validation of the proposed solutions. The ultimate target of the project lies in seamlessly integrating these approximate circuit solutions into cryptographic applications.
Beyond the exciting scientific challenges, this PhD will take place in Grenoble, a picturesque city nestled in the French Alps. The research will be conducted at the CEA LIST site, in collaboration with the TIMA laboratory, surrounded by a vibrant community of experts in embedded software, IoT cybersecurity, hardware design, and machine learning. Together, we aim to redefine the landscape of digital circuitry, contributing to a sustainable future while enjoying the splendid surroundings and collaborative spirit of Grenoble. Join us in pioneering the next frontier of energy-efficient cryptography!