Session

DescriptionAs AI models grow in complexity, the need for optimized architectures and efficient inference mechanisms becomes more pressing. This session explores cutting-edge advancements in the optimization of machine learning models, with a focus on efficiency, scaling, and hardware optimization. The first set of presentations delves into innovative approaches for enhancing the efficiency of edge ML models. The middle presentations shift the spotlight to large-scale models, addressing key challenges in dynamic graph processing, scaling laws for GNNs. Finally, the session concludes with a look at GPU-side optimizations, showcasing techniques for efficient scheduling to meet the demanding requirements of large model serving. This diverse range of topics provides a comprehensive view of the future of ML model optimization from edge devices to high-performance computing systems.