Development of Digital Twin for Liquid Cooled Data Centers
PI: Wangda Zuo
Position Scope
The increasing demands of artificial intelligence and cloud computing are pushing the boundaries of high-performance computing (HPC) infrastructure, resulting in significant growth in data center energy consumption. Data centers now account for approximately 4% of electricity consumption in the U.S., with over 40% of this energy attributed to cooling systems alone. As compute-intensive workloads raise the thermal density of data centers, there is a critical need for advanced thermal management strategies, such as liquid cooling instead of conventional air cooling.
This project seeks to address this urgent challenge by developing a physics-based digital twin of liquid cooling systems for data centers. To capture the complex interaction of thermal-fluid and electric system in the liquid cooling, this project will leverage the equation-based object-oriented Modelica modeling language for high-fidelity simulation of multi-domain systems. The digital twin will serve as the representation of the physical cooling environment to enable the optimal design and control of data center cooling systems.
Specific Objectives
Develop a digital twin model for a liquid-cooled data center
Draft a peer-reviewed journal paper
Long-Term Goal
The long-term objective is to develop and demonstrate a digital twin for liquid cooled data centers. It will serve as preliminary results for a future proposal to the Department of Energy or industry sponsors with a focus on high-performance computing infrastructure.
Plan for Funding
A Ph.D. student at 50% RA in the summer of 2026
Desired Qualifications and Competencies
Specialization in architectural engineering, or mechanical engineering
Post-Comprehensive Ph.D. student
Proficiency in Python with strong coding skills
Familiarity with the Modelica language
Solid background of data center cooling systems
Strong collaboration and teamwork skills
ICDS Engagement
PI Zuo is an ICDS affiliate faculty and has actively engaged with ICDS on its smart city initiatives. This project can contribute to the ICDS community through the organization of research seminars, participation in the ICDS Symposium, and the coordination of a specialized workshop on digital twins for sustainable high-performance computing. These activities are intended to foster interdisciplinary collaboration and advance the field of computational sustainability.