Energy-efficient Embedded Artificial Intelligence: Concepts, Technologies, and Applications

The 2nd edition of the Bordeaux Summer School – Energy-Efficient Embedded Artificial Intelligence (E3AI) builds on the foundations and success of the 2023 inaugural program. While the first edition offered essential training in energy-efficient machine learning, edge computing, and emerging nanoscale technologies, the 2025 edition expands significantly in both scope and technical depth.

Unlike the first edition, which centered primarily on the hardware–software interface for embedded AI, the 2025 program places ferroelectric technologies at the core of energy-efficient, next-generation AI systems. This thematic shift represents the most substantial evolution of the school and reflects the rapidly growing impact of ferroelectric materials and devices in edge intelligence.

The 2025 edition introduces three major advancements:

• A greatly expanded focus on ferroelectric materials and devices, supported by breakthroughs from European initiatives such as FVLLMONTI, FIXIT, Ferro4EdgeAI, and FerroFutures. Ferroelectricity now spans the full curriculum—materials, modelling, circuits, architecture, and system integration.
• A stronger cross-disciplinary framework bridging materials science, device physics, circuit design, compact and TCAD modelling, and system-level co-optimization.
• New hands-on activities and case studies, including compute-in-memory architectures, RF and secure-edge applications, and novel nonvolatile and reconfigurable device technologies.

This expanded program positions the 2025 E3AI Summer School as a leading international training event in future sustainable AI hardware technologies.