AI Data Center Power Demand, 1, 100 TWh Global Use, 30 x US Growth, and Key Energy Deals (2025 to 2035)

AI Data Center Grid Constraints, Power Demand vs. Supply Mismatch

The exponential growth of artificial intelligence is creating a structural deficit in power supply, where data center electricity demand is outpacing the development of new generation and transmission capacity, making grid access the primary constraint on AI expansion.

AI Demand Surge vs. Grid Capacity

The scale of energy consumption is creating a fundamental challenge for utility infrastructure. The following points illustrate the growing mismatch between AI’s power needs and the grid’s ability to deliver.

• Global data center electricity consumption is projected to reach 1, 100 TWh in 2026, an amount comparable to the national consumption of Japan. This rapid increase strains existing electrical grids that were not designed for such concentrated load growth.
• In the United States alone, power demand from AI data centers is forecast to increase more than 30-fold, rising from 4 GW in 2025 to 123 GW by 2035. This surge directly competes with other electrification initiatives for limited grid capacity.
• The inability of electricity supply to keep pace with tech firms’ power demand is already resulting in canceled energy projects, which threatens to slow the expansion of the AI sector. This indicates that energy availability is now a primary limiting factor for AI development.

The High-Voltage Equipment Bottleneck

Beyond raw power generation, the physical hardware required to connect data centers to the grid has become a critical chokepoint. Securing this equipment is now a key factor in project timelines and success.

• The primary bottleneck in deploying gigawatt-scale data centers has shifted from computing hardware to the procurement of high-voltage transmission equipment, particularly transformers. Lead times for these components can delay projects by years.
• A secondary but equally critical bottleneck is emerging in the manufacturing and testing capacity for optical interconnects and silicon photonics. This equipment is essential for high-speed data transfer within and between massive data centers.
• Companies that can secure access to this constrained industrial hardware will be able to execute their growth plans, while competitors who cannot will face significant and costly delays.

$13.6 B Baker Hughes Deal, Energy Sector Investment in AI Infrastructure

Major energy and industrial players are making multi-billion-dollar acquisitions and investments to vertically integrate the supply chain that supports AI infrastructure, signaling a strategic shift to control the physical assets underpinning the digital economy.

Baker Hughes $13.6 B Chart Industries Acquisition

The acquisition of Chart Industries by Baker Hughes exemplifies a strategic move to control the hardware supply chain that is critical for both the energy transition and the AI build-out. This deal provides a blueprint for future market consolidation.

• On July 29, 2025, Baker Hughes announced an all-cash deal valued at $13.6 billion to acquire Chart Industries. This is a definitive statement of intent to dominate the manufacturing of equipment for the entire energy value chain.
• The acquisition combines Baker Hughes’ expertise in energy technology with Chart Industries’ portfolio of equipment for LNG, hydrogen, and carbon capture. This positions the combined entity to supply critical hardware for new power projects designed to meet AI data center demand.

Venture Capital Focus on Grid-Enabling Technologies

Venture capital investment is increasingly flowing towards companies that provide hardware and software solutions to grid constraints. These investments are early indicators of where the market sees future value and potential acquisition targets.

• Antora Energy, a thermal battery startup, announced a $150 million Series B funding round in late 2024. The capital is designated for scaling up its US manufacturing capabilities to address grid stability challenges.
• Startups developing grid-enabling technologies, such as advanced batteries and grid management software, are becoming prime acquisition targets for larger energy and technology firms seeking to de-risk their expansion plans.

Table: Strategic Investments in AI-Adjacent Energy Infrastructure

Vattenfall and Project Enki Offshore AI Data Center Alliance

Strategic partnerships between technology startups, established energy utilities, and industrial giants are forming to develop novel infrastructure solutions that bypass traditional grid constraints, with offshore wind-powered data centers emerging as a leading example.

Project Enki, Vattenfall, and ABB Collaboration

The alliance between Project Enki, Vattenfall, and ABB represents a new model for AI infrastructure development that decouples data centers from terrestrial grid limitations by co-locating them with offshore power generation.

• In June 2026, Swedish energy company Vattenfall partnered with Project Enki and ABB to explore the integration of AI data centers with offshore wind farms across Europe.
• This collaboration aims to build data centers at sea, directly adjacent to wind farms, to utilize power that might otherwise be curtailed. This approach simultaneously addresses power sourcing, transmission, and cooling challenges.

Hyperscaler Power Purchase Agreements

Hyperscale cloud providers like Google, Amazon, and Microsoft have historically relied on large-scale Power Purchase Agreements (PPAs) to meet their renewable energy goals. However, the scale of AI is pushing them toward more direct infrastructure partnerships.

• Large technology firms are among the most aggressive purchasers of renewable energy, but the increasing scarcity of grid connection points is limiting this strategy’s effectiveness for future growth.
• The limitations of the PPA model are driving interest in new approaches, including direct investment in generation and partnerships with energy giants like Next Era Energy to develop dedicated power solutions for data center campuses.

Table: Key Partnerships for AI Power Infrastructure

SWOT Analysis of AI Power Infrastructure Strategies

The strategy of building dedicated power infrastructure for AI data centers possesses the strength of grid independence and long-term cost savings but faces significant weaknesses in upfront capital costs and regulatory hurdles, with the opportunity to monetize wasted energy and the threat of competing grid-enhancement technologies.

Table: SWOT Analysis for Dedicated AI Power Infrastructure

AI Power Scenario, M&A Surge for Electrical Equipment Suppliers

If the current trajectory of AI compute growth continues, expect a significant increase in mergers and acquisitions targeting manufacturers of critical electrical hardware like transformers and switchgear, as tech giants and utilities move to secure their supply chains.

Monitoring M&A and Supply Chain Control

The acquisition of industrial hardware manufacturers is the clearest signal of a strategy focused on controlling the physical world. Observers should monitor M&A activity in this space as a lead indicator of major strategic moves by energy and technology firms.

• The $13.6 billion acquisition of Chart Industries by Baker Hughes serves as a precedent for vertically integrating the supply chain. Watch for similar acquisitions of transformer, switchgear, and cooling system manufacturers.
• Tech giants and major energy players will increasingly view buying their suppliers as a necessary step to de-risk multi-billion-dollar AI expansion plans and ensure project timelines.

The PPA as a Strategic Asset

The ability to secure power will become the defining competitive moat in the AI industry. Long-term, fixed-price Power Purchase Agreements (PPAs) and direct ownership of generation assets will be critical differentiators.

• The value of a company’s AI ambitions will be directly tied to its portfolio of secured power contracts. This will create a new class of energy “haves” and “have-nots” among technology companies.
• Companies that successfully execute a dual strategy of securing near-term power through PPAs while investing in next-generation infrastructure models, such as offshore data centers, will be best positioned to win.

The questions your competitors are already asking

This report covers one angle of the energy constraints on AI growth. The questions that matter most depend on your work.

• High voltage transformer manufacturers and lead times
• Data centers building their own power plants
• Recent tech company power purchase agreements
• US grid interconnection queue wait times

This report does not answer these. Enki Brief Pro does.

Your question, your angle, your framework. SWOT, PESTL, scenario modelling. The same niche depth, built around the decision your work actually depends on.

Run your first brief in Enki Brief Pro