Powering  Intelligence: The Evolving Landscape of Clean Energy & Datacenters

When discussing AI, we can't ignore the rapid growth of data centers and the amount of energy they consume.

Reports show that data center power use has soared since 2020, and analysts expect even more growth over the next five years. By 2030, data centers could represent 6–10% of total energy demand.

AI drives much of this expansion, but it’s part of a larger trend involving digitalization and the electrification of vehicles, buildings, and industry.

Climate Concerns

From a climate viewpoint, the best solution is to power these loads with clean energy. Thankfully, renewables like solar, wind, and battery storage are often the cheapest sources of electricity today. 

That makes them both the environmentally and economically correct choice. When combined with limited fossil fuel-based generation, these renewables can provide the reliable energy data centers need.

Potential Energy Solutions

Despite this, new leaders in D.C. are promoting natural gas as a mostly standalone fix. They’re doing this even though gas plants being planned today won’t be ready until late 2029, at the earliest. 

Meanwhile, “clean firm” technologies, like advanced geothermal and nuclear power, have bipartisan support but are also five to ten years away from large-scale deployment. Some exceptions exist, like Microsoft’s recent 20-year agreement to restart and purchase power produced at Three Mile Island. 

Clean Energy Dominance

In 2024, solar, wind, and battery storage accounted for 90% of all new electricity added to the U.S. grid, with data center operators being the largest buyers of this renewable energy.

According to the U.S. Energy Information Administration (EIA), in 2024:

• Corporations contracted 28 out of 34 gigawatts (GW) of new renewables.

• Amazon, Meta, Google, and Microsoft accounted for 70% of those purchases.

Looking to 2025, the EIA predicts that 93% or 58.4 GW of new capacity will come from solar, wind, and battery storage, while only 7% or 4.4 GW will come from natural gas.

Natural gas faces many challenges, including:

• Geopolitical instability causing fuel price swings

• Closures of aging plants

• Uneven fuel supplies across regions

• Limited pipeline infrastructure

• Rising equipment costs due to shortages

Grid Constraints 

While new renewable energy projections are impressive, they reveal a crucial problem: the grid can’t connect all the new projects waiting in line. Currently, there are about 2,600 GW seeking connection to the U.S. grid, including:

• 1,086 GW of solar

• 1,028 GW of battery storage

• 360 GW of wind

• 79 GW of natural gas

Although energy infrastructure presents a significant upfront expense for data centers, it remains a relatively small portion of their overall operating cost.

Ultimately, data centers require stable electricity and developers are finding innovative ways to provide that power.

Enter Co-Location

An emerging solution is to place energy assets directly next to data centers, avoiding the need to connect to an aging and overstressed grid.

This co-location approach:

• Enables faster deployment while reducing risks of curtailment and transmission delays.
  

• Supports on-site facilities that combine solar, wind, and battery storage to both minimize costs and carbon footprint.
  

• Allows developers to integrate firm power sources such as natural gas to complement renewable resources and ensure 24/7 coverage.

Happening at Scale

This past fall, Intersect Power partnered with Google and TPG and announced plans to complete co-located energy systems in Texas. Their projects, the first of which will be operational in 2026:

• Leverage solar, wind, and batteries to supply approximately 80% of generated power.

• Use natural gas turbines for the remaining 20%.

• Serve as a prime example of how combining renewables with firm power, clean or otherwise, can deliver the reliability industries demand.

Regarding those Growth Projections

While AI and data center growth seems inevitable, long-term power demand projections remain uncertain.

• History reminds us that computing demand forecasts often miss the mark.

• Past experience suggests that efficiency gains, hardware advancements, and new regulations will shape power consumption.

• The industry must stay flexible, combining fast innovation with sound energy planning

The Power Behind Intelligence

AI’s evolution isn’t just a tech challenge, but also an energy challenge.

Meeting its demands requires more than increased power production; it calls for adaptability, vision, and action. Organizations can leverage data centers once considered energy drains as platforms for clean energy integration.

The reality is that clean energy is better for the planet and (more often than not!) the most cost-effective choice. AI can accelerate this shift, turning energy demand into an opportunity for smarter, cheaper, and more resilient power solutions.

***

🛑 Disclaimer: The graphs used in this newsletter are for illustrative purposes and are appropriately credited to the sources provided by the author. GenAI.Works do not claim ownership or copyright and uses them in good faith to share knowledge with our readers.

About the Author

Bill Stark is dedicated to solving climate challenges and is inspired by those leading the way at the intersection of AI and climate. Through this newsletter, he shares his insights, hoping to celebrate pioneers, spark conversations, and inspire more people to contribute to a sustainable future.

🚀 Boost your business with us—advertise where 10M+ AI leaders engage

🌟 Sign up for the first AI Hub in the world.

📲 Our Socials