The College of Computing and Informatics at the University of Sharjah hosted a seminar titled “Resource-Aware Intelligence Across the Computing Continuum,” presented by Dr. Hamza Ouarnoughi from the Department of Computer Science.

The seminar addressed the challenges of deploying artificial intelligence (AI) across edge-to-cloud computing environments. It highlighted the importance of balancing predictive accuracy against resource constraints and operational requirements, including latency, energy consumption, memory usage, communication costs, and hardware limitations.

The session examined approaches to improving AI efficiency, drawing on fields such as machine learning, computer architecture, edge computing, and distributed systems. It also covered performance modeling, hardware-aware neural architecture search, surrogate-assisted optimization, heterogeneous deployment, computation offloading, and federated learning.

In addition, the presentation discussed single-objective and multi-objective optimization techniques, including Pareto-based approaches that help identify trade-offs between competing objectives. These methods support the development of adaptive, resource-aware AI systems that make effective use of available computing resources while meeting varying application requirements.

The seminar concluded by stressing the importance of co-designing AI algorithms and their execution environments to build efficient, adaptive, and deployable intelligent systems across the computing continuum, from edge devices to cloud infrastructure.