machine learning

PowerAtlas: Towards Electricity-Computing Co-Scheduling for Power Systems

arXiv:2607.26710

summary

PowerAtlas is an LLM‑agent framework that jointly schedules electricity supply and computing workloads in data centers, ensuring grid operational constraints and computing service agreements, and is validated on real utility data and a new benchmark (ECBench).

Abstract

The rapid growth of AI workloads is turning data centers into large-scale, volatile, yet spatiotemporally flexible grid loads, creating an urgent need for coordinated electricity-computing scheduling. Under stringent grid constraints, schedules from general-purpose large language models (LLMs) are often infeasible, causing line-flow violations and unserved load. We present PowerAtlas, an LLM-agent framework for electricity-computing co-scheduling that integrates historical instances, domain knowledge, and physical constraints to produce joint decisions satisfying both grid operational rules and the service-level agreements (SLAs) of computing tasks. Working with a provincial power utility in China, we built an experimental electricity-computing network and validated the decision loop on real data-center data; from de-identified operational data we further constructed ECBench, a benchmark of 2,000 scheduling instances with oracle-optimal solutions. Experiments across eleven LLMs demonstrate the effectiveness of PowerAtlas under realistic physical operating conditions, with consistent feasibility and cost gains across three open-weight backbones. Our code is publicly available at https://github.com/JAVA-Jiang/PowerAtlas.

17 pages, 9 figures, 5 tables. Code: https://github.com/JAVA-Jiang/PowerAtlas

Topics & keywords

#electricity-computing co-scheduling#llm agents#grid constraints#data center scheduling#benchmark datasetlarge language modelco-schedulingpower griddata center workloadsECBenchfeasibility optimization