Rema Algunaibet
Physical AI Solutions Architect, NVIDIA
Stefano Mensa
Quantum Developer Relations Manager for Supercomputing Centres and National Laboratories, NVIDIA
Abstract
Modern high-performance computing has evolved far beyond traditional simulation, transforming into a heterogeneous ecosystem where GPU acceleration acts as the foundation for entirely new computational paradigms. This consolidated 3-hour masterclass explores the frontier of these next-generation workflows, where classical GPU-accelerated clusters serve as the orchestrators for advanced intelligence and novel processing units.
The first half focuses on Physical AI, demonstrating how HPC infrastructures leverage GPUs to power real-time video intelligence and vision-language models (VLMs) that bridge the gap between visual observation and robotic action planning. We examine how camera footage is transformed into queryable information, event summaries, and contextual insights to support agentic task planning before physical robotic manipulation begins.
The second half shifts to next-generation co-processing architectures, detailing how these same GPU-dense classical systems are integrated with Quantum Processing Units (QPUs) to support hybrid quantum-classical supercomputing. The discussion will cover the systems architecture linking video analysis, agentic reasoning, and quantum acceleration, highlighting the roles of temporal context, data center integration, and hybrid workflows. Attendees will gain a cohesive framework for how modern classical HPC architectures adapt to manage complex, multi-layered workflows for tomorrow’s autonomous and accelerated systems.
Biographies
Rema Algunaibet is an expert in Physical AI, specializing in video understanding pipelines and agentic reasoning architectures for robotic systems. Her work centers on transforming visual observations into actionable insights, enabling robots to interpret context and plan decisions prior to mechanical execution. She focuses on the practical integration of vision-language models and temporal context within physical AI workflows to bridge high-level reasoning and physical execution.
Stefano Mensa is the NVIDIA Quantum Developer Relations Manager for Supercomputing Centres and National Laboratories, where he works to help NVIDIA partners integrate quantum processors into the world’s leading supercomputing resources in order to build accelerated quantum supercomputers and the applications that will run on them. Previously, Stefano worked at the STFC Hartree Centre, a UK-based national laboratory and supercomputing centre serving as the strategic lead for quantum computing and emerging technologies.
Across his career, Stefano has been committed to enabling the quantum computing community, especially around applications and quantum-classical integration. He’s especially passionate about the convergence of advanced computing technologies and believes useful quantum computing will be driven by datacenter integration and hybrid workflows. Stefano holds a PhD in Theoretical Chemistry from the University of Liverpool, where he is also an Honorary Senior Research Fellow.
Seats are limited
Tutorials run on Monday, 26 October 2026 at KAUST. Attendance is free, but registration is required.
