# Scaling AI Infrastructure Without Water Constraints Transitioning to zero-water thermal architectures to sustain gigawatt-scale AI data center growth. - **Source:** [computerweekly.com](https://www.computerweekly.com) - **Published:** 2026-10-28 - **Topic:** [Business](https://brifing.si/c/business) - **Original article:** - **On Brifing:** https://brifing.si/s/12248f3d-8e4a-499c-b990-5f905178bce1 ## Gigawatt Scale AI Demands Modern AI facilities are rapidly evolving as 100kW-plus racks push infrastructure toward gigawatt-scale requirements. This massive increase in power density creates significant thermal management challenges for operators. Traditional cooling methods often rely heavily on water consumption, which is becoming a critical constraint for site viability in resource-stressed regions globally. ## Water as a Constraint Water availability is no longer just an operational detail but a fundamental limiting factor for data center expansion. As AI workloads intensify, the reliance on water-intensive cooling systems threatens the sustainability and regulatory approval of new sites. Operators must now prioritize thermal strategies that decouple facility growth from local water supply. ## Zero-Water Thermal Architectures John Keane of Johnson Controls advocates for the adoption of zero-water thermal architectures to mitigate these risks. By moving away from evaporative cooling systems, facilities can maintain high-density AI operations without depleting local water resources. This shift is essential for ensuring that infrastructure projects remain viable in diverse climate zones. ## Technical Cooling Shifts The transition involves replacing traditional water-based heat rejection with advanced air-cooled or closed-loop refrigerant systems. These technologies are designed to handle the extreme heat generated by high-density racks while maintaining operational efficiency. Implementing these architectures requires a fundamental redesign of how thermal energy is managed within the data center environment. ## Securing Site Viability Long-term site viability depends on the ability to operate independently of water scarcity issues. By integrating zero-water solutions, companies can secure their expansion plans against future regulatory restrictions and environmental volatility. This strategic move protects capital investments and ensures that AI infrastructure can scale reliably as power demands continue to climb. ## Operational Efficiency Gains Beyond water conservation, zero-water architectures often lead to more predictable thermal performance across the facility. By eliminating the complexities of water treatment and supply chain dependencies, operators can streamline their maintenance protocols. This operational simplicity is a significant advantage when managing the massive, high-density environments required for modern AI model training. ## Strategic Infrastructure Planning Planning for gigawatt-scale AI requires a holistic approach that balances power availability with resource sustainability. Developers must evaluate thermal management early in the site selection process to avoid future bottlenecks. Integrating zero-water technology at the design phase is more cost-effective than retrofitting existing facilities to meet new environmental compliance standards. ## Future of AI Scaling The future of AI growth is inextricably linked to the efficiency of its underlying physical infrastructure. As power densities rise, the industry must innovate beyond traditional cooling paradigms to maintain momentum. Adopting zero-water thermal architectures is a necessary evolution to ensure that AI development remains sustainable, scalable, and environmentally responsible. ## Sources covering this story - https://www.computerweekly.com/opinion/Beyond-power-How-to-avoid-water-as-a-constraint-on-AI-growth — --- _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._