Carbon-Aware Compute--Power Scheduling for AI Data Centers with Microgrid Prosumer Operations

TL;DR

This paper presents a MILP framework for carbon-aware scheduling in geographically distributed AI data centers with microgrid prosumer capabilities, optimizing workload placement, energy use, and grid interaction to reduce emissions.

Contribution

It introduces a joint MILP model that integrates workload scheduling, energy dispatch, and grid interaction, capturing workload flexibility and prosumer operation for sustainable AI data centers.

Findings

Joint MILP improves operational benefit and reduces emissions compared to baselines.

Inference-routing flexibility significantly adds value.

Battery storage and local generation enhance flexibility and sustainability.

Abstract

AI data centers are increasingly becoming tightly coupled compute--energy systems, where workload placement, cooling demand, electricity procurement, storage operation, and carbon emissions interact over time. This paper studies carbon-aware compute--power scheduling for geographically distributed AI data centers with microgrid prosumer capabilities. We propose a mixed-integer linear programming (MILP) framework that jointly schedules rigid training jobs, routes elastic inference workloads, dispatches local generation and battery storage, and manages bidirectional grid interaction under latency, continuity, power-balance, and carbon-budget constraints. The model captures two key features of emerging AI infrastructure: heterogeneous workload flexibility and site-level energy prosumer operation. Experiments on synthetic yet practically motivated instances show that the proposed joint MILP substantially improves total operational benefit over compute-only and energy-only baselines while reducing emissions. The results further indicate that inference-routing flexibility is a major source of value, battery storage provides useful temporal flexibility, and local-generation-rich settings are particularly favorable. The framework provides a tractable optimization abstraction for sustainable and grid-interactive AI data centers.