Ms. Nikhila Pothukuchi

Robust and Trustworthy AI Hardware

Abstract:

Artificial intelligence (AI) is increasingly pervasive in critical applications, necessitating robust and trustworthy systems. While software solutions have been a primary focus, this presentation emphasizes the crucial role of hardware in ensuring AI reliability and security. We explore hardware-level vulnerabilities, including manufacturing defects, environmental effects, hardware Trojans, side-channel attacks, and supply chain risks. We then discuss hardware-based mitigation techniques, such as fault tolerance, robust circuit design, hardware security modules, trusted execution environments, and physical unclonable functions. Finally, we examine emerging trends like neuromorphic and quantum computing, and the potential of AI for designing more robust and secure hardware. This work highlights the importance of a hardware-centric approach to building trustworthy AI systems and identifies key areas for future research and development.