The Shift from Sustainability to Execution

The AI infrastructure boom is colliding with the physical realities of power, cooling, and utility capacity. As workloads scale, the constraints once relegated to facilities management are migrating into the center of IT strategy.

In an interview at HPE Discover 2026, Andrew Desrojects, principal technologist for sustainable transformation at HPE, noted that energy availability, cooling infrastructure, resource utilization, and utility constraints are increasingly steering IT decisions as AI moves from experimentation to production (Data Center Knowledge reports). This represents a fundamental change in how organizations approach deployment.

The era of assuming energy would be available whenever infrastructure was ready is ending. As operators deploy larger AI clusters and increase rack densities, the conversation has shifted from sustainability-focused discussions to operational efficiency. IT teams are no longer delegating grid connection and power requirements to facilities teams; they are actively asking how to consume less energy and extract the maximum amount of IT work from the energy they deploy.

This shift is driven by several physical bottlenecks:

• Utility interconnection timelines.
• Power distribution equipment availability.
• Facility cooling requirements.
• Scrutiny over water consumption tied to new developments.

The implication for leadership is clear: AI investment cannot be viewed solely through the lens of compute power or model performance. If an organization cannot manage the energy and utility constraints, the hardware becomes a stranded asset. The priority has moved from "how much can we build" to "how efficiently can we execute within existing constraints."

As you plan your next infrastructure expansion, ask: Is your IT strategy accounting for the physical limits of your power and cooling, or are you still assuming the grid will simply keep up?