AI data centers are running out of power on Earth, so some of the world's biggest tech companies are looking up. The explosive growth of artificial intelligence has created an unprecedented demand for electricity that existing grids simply cannot meet fast enough. This energy crunch is forcing hyperscalers to explore unconventional solutions, including space-based solar power paired with orbital data centers, a concept that was once purely theoretical but is now attracting serious investment and technical scrutiny.

What Makes Space-Based Solar Different for AI?

Space-based solar has been discussed for decades as a potential energy source, but the traditional vision involved beaming microwave energy back to Earth from massive orbital arrays. That approach faced enormous cost and efficiency barriers. However, recent analyses from Bank of America and Boston Consulting Group suggest a fundamentally different model makes more sense for the AI era: instead of transmitting energy downward, companies would co-locate solar panels with computing infrastructure in orbit, then send processed data back to Earth via optical links.

This shift in thinking reflects a practical reality. Satellites in the right orbit can access near-continuous sunlight, potentially circumventing the land constraints and interconnection queue delays that plague ground-based data centers. In parts of the United States, grid interconnection queues can stretch five to ten years, while dedicated microgrids paired with onsite generation can bring data center capacity online in roughly two to three years.

However, space-based data centers would not be suitable for all workloads. They would struggle with latency-sensitive applications or training large language models, which require fast response times and extremely high densities of both chips and power. Instead, plausible use cases include batch processing of documents, images, and videos; enterprise work; data tagging and classification; and processing satellite-generated data like climate monitoring and disaster-response imagery.

How Much Could Space-Based Solar Actually Cost?

The economics remain daunting. Bank of America estimates that a one-gigawatt orbital data center would cost around $170 billion, roughly three times more expensive than a similarly sized facility on Earth. Despite this premium, the market potential is enormous. Analysts project that if space-based solar and data centers succeed, they could scale global solar cell sales from $60 billion in 2025 to $700 billion by 2035, capture as much as 15 percent of the global AI data center market by 2040, and reach $320 billion in annual revenue.

The path to viability hinges on three major breakthroughs:

• Launch Cost Reduction: Boston Consulting Group estimates that at current launch costs of roughly $1,500 per kilogram, putting one gigawatt of compute capacity in orbit would cost approximately $30 billion in launch expenses alone. Those costs need to drop to around $100 per kilogram to make orbital computing economically plausible.
• Solar Cell Efficiency: Space-based data centers will need to transition from expensive gallium-arsenide solar cells to perovskite-silicon tandem cells, which have more than six times the power-to-mass ratio and offer up to a 15-fold reduction in cost-per-watt. Every kilogram saved on solar panels would yield millions of dollars in launch savings.
• Thermal Management and Radiation Protection: Significant technical changes are required to manage heat dissipation and protect components from radiation in the space environment, adding complexity to system design.

The challenge is compounded by a geopolitical reality: China currently controls over 90 percent of global perovskite-silicon cell manufacturing, including 60 percent of perovskite capacity and 75 percent of perovskite patents. The country is roughly one to two years ahead of the United States in scaling perovskite production.

Who Is Leading the Space-Based Solar Push?

SpaceX and Elon Musk's ecosystem of companies are positioned to dominate this emerging sector. SpaceX has a rare combination of capabilities: reusable launch technology, satellite manufacturing, an existing optical communications network via Starlink, and close ties to AI chip and model development. Earlier this year, SpaceX filed with the Federal Communications Commission for permission to deploy up to one million satellites as part of an orbital data center constellation called "Starmind," designed to carry up to 250 kilowatts of peak computing power.

Musk estimates that SpaceX could eventually deploy up to 500 gigawatts per year of solar-powered AI satellites, roughly equal to around 60 percent of average U.S. electricity demand. While that projection is ambitious, it underscores the scale of ambition driving the space-based solar vision.

The U.S. advantage lies in launch economics. SpaceX's launch costs are around $1,500 per kilogram, compared to approximately $7,000 per kilogram for even the cheapest Chinese launches. While SpaceX continues to reuse its boosters, Chinese companies are still testing recovery and reuse systems.

What's Happening on the Ground While We Wait for Space?

While space-based solutions remain years away, hyperscalers are pursuing more immediate alternatives to address the power crunch. On September 23, 2026, Zeo Energy and Ewyze announced a partnership to develop integrated off-grid power and data center infrastructure across the United States. Ewyze's model pairs solar generation and battery storage with a dispatchable power source, which may be a grid connection, natural gas, or geothermal energy, to build dedicated microgrids alongside data center campuses.

"The artificial intelligence revolution is fundamentally a power challenge," said Svante Kumlin, CEO of Ewyze. "AI workloads demand vast amounts of reliable, cost-effective energy, and the existing grid simply cannot deliver it at the speed or scale the market requires."

Svante Kumlin, CEO at Ewyze

This ground-based approach reflects a broader trend. Rather than waiting for breakthrough technologies, companies are integrating renewable generation, battery storage, and data center assets into single, coherent platforms. Battery Energy Storage Systems (BESS) are increasingly being positioned as part of the power architecture rather than simply as backup equipment, supporting critical backup power, ride-through capability, peak shaving, renewable firming, black-start capabilities, and grid services.

The AI Data Center BESS market reached $1.65 billion in 2025 and is projected to reach $31.97 billion by 2035, expanding at a 34.5 percent compound annual growth rate. This explosive growth reflects the urgency of the power challenge. North America currently represents the largest regional market, while Asia-Pacific is expected to record the fastest growth.

How to Evaluate Data Center Power Solutions

• Grid Interconnection Timeline: Assess whether traditional grid connection is feasible within your project timeline. If interconnection queues exceed five years, dedicated microgrids or onsite generation become more attractive alternatives that can reduce deployment time to two to three years.
• Power Density Requirements: Determine whether your workload is latency-sensitive or can tolerate delays. Space-based and off-grid solutions work best for batch processing, enterprise work, and data classification, not for real-time AI model training or inference.
• Long-Term Cost Structure: Compare the total cost of ownership across ground-based microgrids, traditional grid connection, and emerging technologies. While space-based solar remains expensive today, launch cost reductions could make it competitive within a decade.
• Energy Storage Integration: Evaluate battery storage systems as part of your power architecture. BESS can provide resilience, manage peak demand, support renewable integration, and enable grid services that improve overall system economics.

The reality is that no single solution will dominate. Ground-based microgrids with renewable generation and battery storage are solving the immediate power crisis for hyperscalers. Meanwhile, space-based solar represents a longer-term bet on breakthrough economics and technology maturation. Both approaches reflect the same underlying truth: AI's power appetite has fundamentally reshaped how we think about energy infrastructure, forcing innovation across multiple fronts simultaneously.