![[2511.18554] Online Smoothed Demand Management](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2511.18554] Online Smoothed Demand Management
Summary
The paper introduces Online Smoothed Demand Management (OSDM), a framework for optimizing energy purchasing and delivery in data centers, balancing base and flexible demand while minimizing costs and promoting grid stability.
Why It Matters
As energy consumption in data centers rises, effective demand management becomes crucial for sustainability. OSDM addresses this by providing a structured approach to energy management that can enhance operational efficiency and reduce costs, making it relevant for energy-intensive industries.
Key Takeaways
- OSDM optimizes energy management for data centers by balancing demand types.
- The proposed PAAD algorithm achieves an optimal competitive ratio in energy purchasing decisions.
- A novel learning framework enhances performance by adapting to historical data.
- The study highlights the importance of smooth energy delivery to promote grid health.
- Case studies demonstrate significant performance improvements with the proposed methods.
Computer Science > Data Structures and Algorithms arXiv:2511.18554 (cs) [Submitted on 23 Nov 2025 (v1), last revised 20 Feb 2026 (this version, v2)] Title:Online Smoothed Demand Management Authors:Adam Lechowicz, Nicolas Christianson, Mohammad Hajiesmaili, Adam Wierman, Prashant Shenoy View a PDF of the paper titled Online Smoothed Demand Management, by Adam Lechowicz and Nicolas Christianson and Mohammad Hajiesmaili and Adam Wierman and Prashant Shenoy View PDF HTML (experimental) Abstract:We introduce and study a class of online problems called online smoothed demand management $(\texttt{OSDM})$, motivated by paradigm shifts in grid integration and energy storage for large energy consumers such as data centers. In $\texttt{OSDM}$, an operator makes two decisions at each time step: an amount of energy to be purchased, and an amount of energy to be delivered (i.e., used for computation). The difference between these decisions charges (or discharges) the operator's energy storage (e.g., a battery). Two types of demand arrive online: base demand, which must be covered at the current time, and flexible demand, which can be satisfied at any time before a demand-specific deadline $\Delta_t$. The operator's goal is to minimize a cost (subject to above constraints) that combines a cost of purchasing energy, a cost for delivering energy (if applicable), and smoothness penalties on the purchasing and delivery rates to discourage fluctuations and encourage ``grid healthy'' decisions...
