Megawatt Framework Turns Battery Storage into AI Lever
Get the Energy newsletter
Daily energy & climate — solar, EVs, oil, the policy fights and tech bets shaping the transition. Free.
- Compute Per Megawatt (CPM) is a new framework emerging for hyperscalers and large load users to leverage behind‑the‑meter battery storage for AI compute as power demand rises.
- Energy storage projects traditionally aim to earn capacity payments, build resilience, and participate in demand‑response programs.
- Battery storage is reframed from a revenue‑generating asset to a lever that expands the IT power budget, enabling higher compute utilization despite grid constraints.
- Alejandro de Diego of Modo Energy explained that storage does not make GPUs more efficient but removes constraints from power capping and workload scheduling, effectively expanding the compute budget.
- Battery Business & Development Forum co‑organized by pv magazine will discuss utility‑scale storage on March 31 in Frankfurt.
Why it matters: Hyperscalers can now tap additional compute capacity without being limited by grid constraints, turning battery assets into performance levers; meanwhile, the focus shifts from traditional capacity‑payment revenue to directly supporting AI workloads. This reframing also highlights the strategic value of behind‑the‑meter storage for data‑center resilience and demand‑response participation.
Ask SkimNews
