# NTT Global Data Centers Analyzes AI Infrastructure Deployment Constraints

Scaling AI capacity requires navigating complex permitting processes and community misconceptions regarding resource consumption.

- **Source:** [www.theregister.com - Articles](https://brifing.si/source/2c34ea50-d341-4a10-925d-12c012a9ad10)
- **Published:** 2026-10-16
- **Topic:** [Business](https://brifing.si/c/business)
- **Original article:** <https://www.theregister.com/ai-and-ml/2026/10/16/sponsored-global-data-centers-weighs-the-price-tag-of-not-building-fast-enough/5303402>
- **On Brifing:** https://brifing.si/s/b58fe116-8444-444e-b4f6-7695129829c1

## The Scale of AI Demand

Modern data centers are evolving into massive facilities covering multiple football fields. NTT Global Data Centers reports that these sites now require hundreds of megawatts, far exceeding the ten-megawatt standard of the past decade. This shift creates a direct collision between the rapid growth of AI-driven demand and physical infrastructure realities.

## Financial Projections for Infrastructure

NTT’s AI Stress Test Report models three distinct growth trajectories leading up to 2030. A baseline scenario projects 23 percent annual demand growth, necessitating approximately $3.8 trillion in digital infrastructure investment. If growth accelerates, this capital requirement climbs to $5.7 trillion, highlighting the massive scale of required global investment.

## Permitting as the Primary Bottleneck

While industry focus often centers on GPU availability and switchgear, NTT identifies paperwork and planning as the true gating factors. Bruno Berti, CTO at NTT, emphasizes that hardware speed is irrelevant without the necessary permits. Deployment timelines and power availability currently represent the most significant constraints for modern data center operators.

## Strategic Shift in Expansion

Berti argues against a reactive 'build, baby, build' mentality, which often leads to project stalls. Instead, he advocates for a portfolio approach that views capacity expansion as a system-wide coordination challenge. This method prioritizes strategic planning over a simple land race, ensuring that infrastructure development aligns with local regulatory and physical capacities.

## Addressing Community Water Concerns

Community consultations frequently cite water waste as a major objection to new data center projects. However, modern designs utilize advanced liquid cooling technology to remove heat more efficiently. This technological shift effectively eliminates the excess water waste associated with older facility generations, addressing a primary concern raised during local town hall meetings.

## Grid Integration and Power Stability

Misconceptions regarding power theft from the grid persist, yet operators are increasingly partnering with utilities years in advance. New hundred-megawatt-plus sites often incorporate on-site solar or wind generation. These renewable energy sources help stabilize the local grid rather than draining it, transforming data centers into active participants in regional energy management.

## The Necessity of Transparency

NTT emphasizes that the industry must move away from secrecy to secure necessary permits. Building trust through education and active partnership with local stakeholders is essential for long-term success. Transparency regarding resource usage and grid benefits serves as a foundational element for gaining community support throughout the current decade of expansion.

## Future Outlook for Operators

The ability to turn capacity into reliable AI infrastructure depends on navigating these complex socio-technical constraints. By addressing misconceptions and streamlining permitting processes, operators can better align their growth with the needs of policymakers and communities. Success through 2030 requires a balanced approach that integrates technical innovation with proactive public engagement.

## Sources covering this story

- https://www.theregister.com/ai-and-ml/2026/10/16/sponsored-global-data-centers-weighs-the-price-tag-of-not-building-fast-enough/5303402 — 

---

_Summary written by Brifing from the original reporting. Copyright in the original article belongs to its publisher; read the full piece at the link above._