T
he exponential demand for AI computing capacity has triggered a boom in the data center market, with existing facilities undergoing extensive refactoring and new ones being built to meet the growing need. This trend is global, with major players like the US, China, and Europe re-evaluating their computing footprint to gain an edge in the AI race.
As a result, the entire data center supply chain is experiencing ripple effects, including real estate market fluctuations. Hyperscalers are using their dominant power to secure prime locations and potentially own the entire value chain, from power generation to infrastructure development. Meanwhile, second-tier sites are also benefiting from the demand, leading to a reevaluation of real estate assets and how they're priced and financed.
Data centers have evolved into strategic infrastructure, with metrics shifting beyond traditional square footage and occupancy rates. The emergence of AI workloads has introduced new requirements, including:
* Strategic access to energy grids or electricity production sites
* Proximity to fiber and advanced communication hubs
* Availability of GPUs and GPU acceleration techniques
* Energy efficiency and cooling systems
* Compliance with domestic and international regulatory requirements
Compute, network, or storage demands from AI workloads can require significantly more energy than traditional computing requests. For example, AI Deep Search compute and energy costs are an order of magnitude higher than for traditional Google Search queries.
Energy stability, reliability, and resilience have become the primary factors in evaluating data center facilities. Traditional locations like Virginia, Silicon Valley, and Texas are being reevaluated due to constraints such as natural disasters or high density. Infrastructure investors are turning to second-tier locations that offer more flexibility and economic viability.
GPUs, dominated by NVIDIA, are essential for AI workloads but expensive, backlogged, and subject to strategic interests. Deploying GPUs in concentrated environments can lead to overheating issues, making new generation cooling technologies necessary for dense AI data center facilities.
The cloud has centralized computing, but users have not moved their applications to the cloud due to network limitations. AI workloads with direct user interaction or inference models deployed on edge systems are sensitive to latency and require high-speed fiber connectivity.
Regulatory pressures and sovereignty battles are driving the selection of sites for data center facilities. Governments are increasing their interest in AI infrastructure, encouraging private sector investments. Hyperscalers like AWS, Google, and Microsoft are seeking full ownership of the entire data center supply chain through investments and acquisitions.
Energy companies are developing renewable energy alternatives for data centers, while REITs and infrastructure funds are creating new types of funds adapted to AI-ready data center assets. The strategic value of data centers is shifting, with access to AI-ready facilities becoming a key prize in the digital economy. Traditional real estate metrics need to be extended to reflect the elevated nature of such facilities, including power, GPUs, design, communication, and sovereignty.
