REU Site: Applying Data Science on Energy-efficient Cluster Systems and Applications

Energy efficiency has become a critical challenge in the surge of data-intensive and compute-intensive applications and workloads in modern data centers. Developing effective energy conservation techniques for cluster systems and applications is increasingly important in business planning and decision-making. There is a lack of holistic energy-efficient solutions capable of reducing energy costs by comprehensively leveraging a variety of measurement methods, energy consumption models, and energy-saving strategies within parallel and distributed computing environments. This Research Experience for Undergraduates (REU) site will research and develop energy-efficient solutions for cluster systems and applications. These solutions can serve as guidance for analyzing energy consumption data, building predictive models to estimate the energy cost of cluster systems and applications, and enabling energy-efficient workload management for cluster systems, as well as providing energy-saving solutions for applications. This project will promote engagement of undergraduate students in data science and parallel and distributed computing research, and demonstrate how energy-efficient solutions can be used for social challenges. Research results from the project will be incorporated into university courses to prepare a competitive workforce by teaching high-demand skills. The goal of the project is to increase the participation of undergraduate students in research that integrates data science and parallel and distributed computing for building energy-efficient solutions for cluster systems and applications. The REU site will recruit 10 undergraduate students each year to participate in energy-efficient research by conducting insightful analysis of energy consumption of cluster systems and applications, and investigating energy cost models and energy-saving strategies. REU participants will develop skills in intelligent data collection, data processing, and data visualization. They will gain expertise in applying data science techniques to model the energy consumption of cluster systems and applications. They will also study key concepts of parallel and distributed computing and explore solutions to reduce the energy cost of cluster systems and applications. This project will provide students with an immersive experience through various training and research activities, enabling them to apply data science and parallel and distributed computing knowledge to solve real-world problems. It will also foster communication skills through collaboration with mentors and peers as well as participation in seminar and poster presentations. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.