Power progress in your global data center expansion

The race for data center capacity that can support artificial intelligence (AI) has contributed to a perfect storm: colocation vacancy rates below 3%, rents surging 11% annually and new facilities 72% pre-leased before breaking ground. Power availability timelines can stretch up to a decade in some regions as hyperscalers snap up entire buildings before enterprise customers can act.

For global portfolio managers, this isn't just a capacity crisis—it's also a fundamental shift that requires a new playbook for securing infrastructure, managing power constraints and delivering reliable digital services across borders.

In discussions about global data center power challenges, significant regional variations exist. In US markets, companies face power delivery wait times of 2-3 years in the Mountain West and New York metro areas, but a whopping 8-10 years in the Pacific Northwest. In Northern Virginia, you'll wait three to five years unless your site sits within half a mile of existing power lines.

Similar problems exist worldwide. Singapore's power exhaustion is forcing growth to shift to neighboring countries like Malaysia, with over 1 gigawatt (GW) of new builds completing within a year, and Thailand, where Google is developing three campus sites totaling 540 megawatts (MW). Indonesia is also worth noting, as they are commencing a new phase of data center development. Japan faces similar power availability challenges, while Hong Kong has seen less data center investment due to geopolitical concerns. These power shortages are a universal headache across all global regions, illustrating that power acquisition is a fundamental challenge whether you're operating in North America, Europe or Asia-Pacific markets.

As hyperscalers secure entire buildings to meet their expanding needs, other businesses are adapting by planning more than three years ahead or considering building their own facilities. The industry is also witnessing more strategic partnerships forming between energy providers and data center players—such as:

• Google's collaboration with Engie for their Hanau facility
• Virtus partnering with Germany's national electricity network manager for their new Germany Wustermark Campus—representing a significant evolution in the market
• Echelon Data Centres' €2bn joint venture with renewable power giant Iberdrola to develop data centers across Spain, with Iberdrola providing 100% of power infrastructure and clean energy supply
• CyrusOne KEP, a strategic joint venture between CyrusOne and Japanese utility giant KEPCO, breaking ground on a 48MW hyperscale facility in Japan's Kansai region—with ambitious plans to develop 600MW IT capacity over the next decade
   

While AI has changed what companies need from data centers, it's just part of the story. According to JLL's 2025 Global Data Center Outlook, traditional workloads like data storage and cloud applications still make up most of the demand. Even in the most optimistic scenarios, AI workloads might only reach 50% of data center capacity by 2030. Companies initially used remote facilities for AI training but are now splitting operations between training centers (near power sources) and inference centers (closer to cities for faster response). These AI systems need substantially more power—regular setups use about 12kW per rack while AI systems need between 41-130kW—pushing data centers to develop new cooling and power solutions.

Organizations with global real estate portfolios must plan their space needs in advance rather than waiting until problems arise. The London Stock Exchange Group (LSEG) is a good example of this, having moved from short-term planning to long-term planning across its global portfolio. This shift requires companies to balance owned versus colocation facilities, with the foresight to plan ahead to reserve space in areas where it's limited.

Proximity to power has become a key business decision, not just a facilities issue. Portfolio managers need to create plans for each region based on how long utilities take to connect power, which impacts both the timeline for new space occupation and decisions about delaying exits from existing facilities. When power availability is limited, portfolio strategies differ based on facility type. For owned sites, alternative options like smaller independent power systems or renewable energy sources might be considered. For colocation space, the focus shifts to selecting agile operators who can effectively manage these power planning challenges on behalf of their clients.

The cost of power differs greatly depending on location, so careful financial analysis is required to make the best decisions about where to place operations worldwide.

Data centers have undergone more transformation in the past few years than in the previous two decades, creating the need for enhanced partnerships, planning processes, sophisticated business models, and talent strategies. This acceleration means you’ll need to be more flexible and agile in your planning while at the same time looking further into the future than ever before.

Leading companies now build strong partnerships with key suppliers to manage their worldwide properties effectively. For instance, Compass Data Centers joined forces with Schneider Electric in a $3 billion deal that shows the level of teamwork needed to scale effectively. Compass now uses a manufacturing approach with standard components, which helps them maintain quality across different regions while achieving sustainability goals.

Partners must adjust to local differences while following global guidelines. Andrew Green, who leads JLL's data center team in Asia Pacific, explains that working across 17 countries with different cultures, laws and 15 languages requires partnerships that can adapt to local conditions while still delivering reliable results.

Companies must now actively engage with communities instead of treating this as an afterthought. When local groups show concern over new data center projects in their communities, portfolio managers need to create clear communication plans that address specific regional concerns – such as electromagnetic field worries in Asia or resource consumption in the United States. They must also articulate convincing economic stories that show local stakeholders how projects can benefit their communities.

To achieve excellence across global properties, companies must balance standardized approaches with necessary local adaptations. It’s important to employ advanced management systems to create global standards while adapting to each region's culture and rules. Teams now rely on technology platforms that allow unified control across different languages, laws and time zones—tools that are now considered essential, not just nice to have.

The lack of skilled workers continues to create a major challenge for the industry. Some companies now develop their own training programs to teach new skills as equipment becomes more reliable and job requirements change. Leaders must use management styles and make investments to enforce global standards while giving local teams some independence when supervising staff across different time zones and cultures.

Companies must make their technology adaptable to protect their global properties for the future. You’ll need to design facilities that can handle new artificial intelligence needs while still supporting traditional business systems. Many organizations find this balance difficult because their existing data centers aren't built to the modern design specs needed for AI systems.

Financial stability requires close attention to changing lease costs. Credit ratings now strongly affect pricing, with some markets charging non-investment grade tenants 10% more. Price increase clauses typically start at 3% minimum and renewal terms now often pass tax and insurance costs to tenants. Portfolio managers need detailed cost models that account for these changing expenses across different currencies and economic regions.

The rapid growth of AI technology combined with limited infrastructure has created new management requirements for global data center properties. Successful companies now plan three or more years ahead, build strong relationships with suppliers and communities and perform well across different global settings.

The next one to two years will test portfolio managers as AI systems multiply and power limitations intensify. Leaders who tackle these challenges early—securing space before they need it, building adaptable infrastructure for changing AI needs and creating strong operating systems that work in multiple countries—will help their organizations succeed in the AI-focused future. As the industry changes faster than ever, the ability to combine long-term planning with day-to-day execution has never been more valuable.

Need expert guidance to develop your data center strategy as AI rapidly transforms the industry? Explore our comprehensive data center solutions to help you secure critical power resources, strengthen your property portfolio and build digital infrastructure that will serve your needs now and in the future.