Postdoc, Hewlett Packard Enterprise
2 papers at NeurIPS 2025
DCcluster-Opt introduces a high-fidelity benchmark for sustainable workload scheduling in geo-distributed data centers, integrating real-world data and physics-based models to evaluate multi-objective optimization strategies via a Gym environment.
LC-Opt is a high-fidelity RL benchmarking environment for optimizing liquid cooling in HPC systems, enabling scalable, explainable, and energy-efficient control with RL and LLMs, while promoting optimal data center solutions.
