Hyperscalers also dominate geographically. In a separate Synergy study revealing the world’s top 20 hyperscale data center locations, just 20 U.S. state or metro markets account for 62% of the world’s hyperscale capacity.
Their rapid expansion is driven by AI workloads that behave differently from traditional computing. GPU clusters can spike power consumption to 15 times idle levels in milliseconds, putting unprecedented stress on data-center infrastructure and the grid.
AI’s power needs are growing exponentially. Today’s ~5 GW of AI load could exceed 50 GW by 2030, , according to the report, "Scaling Intelligence: The Exponential Growth of AI's Power Needs," from the Electric Power Research Institute (EPRI). Training frontier models already require 100-150 MW per run, with requirements doubling or tripling each year. By decade’s end, AI could represent more than 5% of total U.S. power generation capacity.
To understand the influence of GenAI, it’s important to understand this: While traditional hyperscale builds were optimized for cost per watt and maximum uptime, new AI workloads demand higher rack densities, liquid cooling compatibility, and flexible compute orchestration.
And the megaprojects being built by big tech firms like Apple, Microsoft, AWS, and Google, provide important signals for the wider data center ecosystem.
- Colo providers are increasingly aligning with hyperscalers through campus-adjacent footprints or serving as short-term capacity offload while long-term builds are underway.
- Power developers (including natural gas and small modular reactor players) are embedding earlier in project planning, often through joint ventures.
- Transmission infrastructure and substation delivery timelines are now pacing part of factors considered in site selection, often more than fiber or land cost.
Energy Transition Comes to the Data Center
Data centers are driving the renewable energy transition as operators and utilities pivot toward new energy and infrastructure models to keep up with demand, including the following:
- Renewables and alternative energy: Wind, solar, geothermal, biofuels, nuclear, and fusion are becoming part of long-range planning. According to the International Energy Agency (IEA), renewables such as wind and solar supplied about 24% of their electricity, while nuclear power supplied about 20% and coal around 15%.
- On-site generation: As grid constraints tighten, data centers are adopting local power sources, including natural gas and small modular reactors. The IEA reports that as of 2024, natural gas supplied more than 40% of electricity for U.S. data centers.
- Next-generation cooling and water management: New systems cut water use and energy waste in high-density AI deployments.
- Battery Energy Storage Systems (BESSs): Replacing diesel generators for backup power and offering grid-support revenue opportunities.
- Waste-heat reuse: Redirecting server heat to nearby buildings or energy systems.
New markets such as Idaho, Louisiana, Oklahoma, and Texas are emerging as operators seek regions with faster timelines, lower costs, and more flexible permitting.
Strategic Imperative: Rethink Compute
Traditional hardware can’t power AI’s potential. That’s because the energy demands of classical chips are unsustainable, driving up costs and using more water for cooling. For every calculation a classical chip makes, it takes inputs, generates outputs, and then releases—as heat—the inputs it no longer needs.
This is driving interest in fundamentally new compute architectures such as reversible computing, where chips can recover and reuse most of the energy they expend. These chips could redefine the economics of AI processing. Companies including Intel, AMD, NVIDIA, AWS, Microsoft, PsiQuantum, and Vaire are investing in this shift.
Bottom Line for the C-Suite
The speed of AI adoption is now directly constrained by the availability, cost, and source of electricity. Over the next decade, power strategy will become as critical to enterprise competitiveness as cloud strategy has been over the past one.
IT leaders and other senior executives who align energy planning, sustainability investments, and AI infrastructure strategy will be best positioned to scale while managing cost, risk, and reputation.
Quiz