Advanced Data Center Power Management (ADCPM)
Mitigating Grid Impacts from AI Power Demand Growth

Data Center operators are under pressure to deliver reliable, high-performance, and sustainable services in a time sensitive manner under severe constraints: power availability from the local utility to meet exponentially growing demand, supply chain limitations, and workforce/skillset availability. Meeting the extraordinary demands of customers, communities, and local governments necessitates the strategic implementation of energy efficiency, renewable integration, cooling innovation, and operational resilience — all while navigating regulatory and competitive landscapes.

Developed with the benefit of ongoing interactions with some of the largest data center users, ADCPM transforms data centers from passive consumers into active participants in the energy ecosystem, capable of forecasting demand, coordinating energy assets, and responding dynamically to grid and market conditions. It extends traditional data center power monitoring into a grid-interactive, forecast-driven control architecture. The following represents key components of an integrated ADCPM architecture.

Inside the Fence

• Intelligent PDUs and branch circuits
• UPS, switchgear, transformers
• DCIM / BMS / SCADA systems
• Real-time load and power quality telemetry

Co-Located Energy Assets

• Battery Energy Storage Systems (ESS)
• Onsite renewables (solar, wind, hydro)
• Backup generation, with SMRs (Small Modular Reactor) as a future option

At the Grid Edge

• Grid constraint awareness (capacity, voltage, congestion)
• High-resolution load forecasting (AI workload–aware)

• Dispatch and scheduling of storage and generation
• Price, carbon, and reliability optimization

Across the Portfolio

• Multi-site workload and power coordination
• Grid-aware compute placement
• Portfolio-level resilience and optimization

OpenDSO™, a distributed IoT platform for monitoring and managing energy infrastructure, is the underlying technology for OES' advanced data center power management concept. Featuring a layered, three-tier architecture, OpenDSO combines grid-edge integration, asset & system management services, and a growing application portfolio capable of monitoring and managing energy infrastructure on both sides of the interconnection. With ADCPM, it acts as the orchestration layer that unifies facility systems, co-located energy assets, and grid interfaces into a single observable and operational domain.

Traditional DCPM measures efficiency. ADCPM with OpenDSO dynamically shapes demand, improves monetization of on-site energy infrastructure, more effectively operates the interconnection, and enables acceleration of growth. The following table shows how OpenDSO manifests that vision: expanding boundaries and capabilities of existing power management for data centers.

An example of OpenDSO in action involves load shifting across multiple sites. Local price spikes or supply constraints may require dynamic workload shifting across data center sites.

OpenDSO can deploy and launch distributed applications that monitor data center PDU load at each of the sites and use AI/ML + simulations to forecast compute and load. The applications allow data center operators to develop and execute an optimal schedule for the combined compute load and executes control actions to shift compute from one data center to another.

Value is provided by reducing the load at the data center(s) with higher unit operating costs, reducing the overall carbon footprint and/or improving overall efficiency.

ADCPM offers data center operators the potential to generate significantly higher value across financial, operational, sustainability, and strategic metrics.

Financial Impact

• Reduce demand charges through forecast-driven peak shaving
• Defer electrical and interconnection upgrades by shaping load profiles
• Improve asset utilization of ESS and generation
• Lower marginal cost of AI-compute via grid-aware workload placement

Operational Resilience

• Faster scaling than grid-only capacity additions
• Reduce exposure to local grid congestion and outages
• Increase predictability of power availability for AI training cycles

Sustainability & ESG

• Higher renewable penetration without reliability risk
• Carbon-aware dispatch and auditable reporting
• Alignment with net-zero and regulatory commitments

Strategic Advantage

• Operate as grid-interactive energy platforms, not fixed loads
• Leverage power as a controllable variable, instead of a growth constraint

OpenDSO… Local Autonomy | Grid-Aware Coordination | Execution at the Edge™