Hi! I am Lorenzo Casalino, a post-doctoral researcher at CentralSupélec (Rennes, France), and integrated in the SUSHI team (INRIA/Irisa). Together with the other members of the ANR ATTILA project, I am exploring the application of the approximate computing paradigm to protect deep neural network implementations against passive side-channel attacks.
I did my Ph.D., issued by the Sorbonne Université (Paris, France), at the CEA-List (Grenoble, France), where I investigated the impact of CPU microarchitectures on the practical security of masked software implementations of cryptographic primitives.
My research interests concern with the hardware security of sensitive applications, the investigation of innovative countermeasures and the employment of tools (for instance, compilation tools) to aid their application.
L. Casalino, M. M. Real, J.C. Prévotet, R. Salvador (2026). “Double Strike: Breaking Approximation-based Side-Channel Countermeasures for DNNs”. IEEE HOST.
A. Plin, L. Casalino, T. Rokicki, R. Salvador (2026). “Knock-Knock: Black-Box, Platform-Agnostic DRAM Address-Mapping Reverse Engineering”. uASC’26 - 2nd Microarchitecture Security Conference.
L. Casalino (2024). “(On) The Impact of the Micro-architecture on Countermeasures against Side-Channel Attacks”. Sorbonne Universite.
L. Casalino, N. Belleville, D. Couroussé and K. Heydemann (2023). “A Tale of Resilience: On the Practical Security of Masked Software Implementations”. IEEE Access.
ATTILA – Addressing securiTy Threats to artIficiaL intelligence in Approximate computing systems. Workshop interdisciplinaire sur la sécurité globale. Paris, France. Mar. 26-27 2025.
A Tale of Resilience: On the Practical Security of Masked Software Implementations. IACR CHES. Prague, Czech Republic, Sept. 10-14, 2023.
“Side Channels and Deep Neural Network Weights: Attacks, Defences and the Future to Come”. Journées SSLR. Paris, France. Oct. 8-9 2025.