Every time you ask ChatGPT a question or stream a show on Netflix, you're using a data center somewhere in America.
These massive warehouses filled with servers need one thing above all else: Lots of power.
A single AI data center consumes enough electricity to power 80,000 homes. As AI adoption explodes across industries, companies are building hundreds of these facilities right now.
There's just one problem - the U.S. power grid can't keep up.
That lack of energy is creating a potential opportunity for investors to profit from the companies aiming to solve AI’s power crisis with new data center energy solutions.
Keep reading to learn how our market analysts view this opportunity, what types of energy are seeing investment, and which stocks are winning the data center energy race.
Power costs are climbing, and so is the price of everything else you buy. Our CEO Jaspreet Singh connects the two on September 29th in his free workshop on how to profit from a dollar that is losing its value.
Where Do Data Centers Get Their Power?
Until recently, most data centers traditionally relied on connections to the electrical grid. That means the electricity came from power plants.
But that was in an era before AI… when power requirements were lower.
But AI changed everything. The computing power needed to train AI models and process calculations requires exponentially more electricity than traditional cloud computing.
Here’s the problem: These AI data centers need to scale in order to meet demand.
Connecting a new AI data center to the grid can take years. Which means these data centers need more energy now.
Many new forms of energy like nuclear, wind, and solar can’t just plug into data centers. Infrastructure needs to be upgraded, first.
Imagine connecting a fax machine to your smartphone - It’s not a practical solution.
Companies like Amazon, Microsoft, and Google can't wait years. They have millions of customers who expect 24/7 uptime.
The solution? On-site and near-site power generation that bypasses the grid entirely.
Natural gas turbines provide this solution today. Unlike grid connections that take years to establish, gas turbines can be installed on-site or nearby and start generating power within months.
The U.S. Department of Energy projects data center energy consumption will double or potentially triple by 2028. Grid capacity simply won't expand fast enough to meet this demand.
This infrastructure gap is forcing data center operators to take power generation into their own hands.
Why the U.S. Is Building So Many Data Centers
The AI boom means the U.S. needs more data centers in order to keep up with demand.
The demand is coming from:
- Cloud computing giants, who need AI infrastructure.
- Businesses who adopt AI tools.
- Consumers using AI chatbots.
- Researchers training AI models.
All of that computing happens in data centers.
The hyperscalers - Amazon, Microsoft, Google, and Meta - are racing to build capacity. Their AI and cloud computing businesses generate billions in annual revenue, but only if they stay reliable.
Amazon Web Services alone generated over $30 billion in revenue in Q2 2025, with $10 billion in operating profit.
Every new AI application requires more computing power.
More computing power means more servers.
More servers mean more data centers.
And more data centers = more power consumption.
The U.S. has a large population with dense urban centers, particularly on the East coast. Data centers are being built near these areas to reduce latency.
But building near cities means competing for limited grid capacity with residential and commercial users.
This is where natural gas turbines create an investment opportunity.
Are Data Centers Using Natural Gas?
The short answer is yes - and the numbers are growing.
Data center operators are signing long-term contracts with gas turbine manufacturers right now. These aren't experimental pilot programs - they're multi-year commitments worth billions of dollars.
The gas turbine market for power generation is projected to grow from $16.7 billion in 2023 to $30.4 billion by 2034, representing a 5.5% compound annual growth rate.
Much of this growth is being driven by data center demand.
Natural gas turbines solve several problems simultaneously:
Speed of deployment: Turbines can be installed and operational within months, not years.
Reliability: They provide continuous power with minimal downtime, unlike intermittent renewable sources.
Flexibility: Power output adjusts to meet demand in real-time.
Regulatory clarity: Gas turbines have decades of established regulations, making permitting faster than experimental technologies.
Companies building data centers need power today. Nuclear solutions might work eventually and space-based data centers are fun to think about, but are largely theoretical.
Renewable energy sources like solar energy aren't as reliable as you can’t turn up the sun, or generate more wind when you need more energy.
Gas turbines work now, which is why hyperscalers are committing billions to this technology.
What Type of Energy Do Data Centers Use?
Data centers have historically used a mix of:
- Coal.
- Natural gas.
- Nuclear.
- Renewables like solar and wind (depending on the region).
But primarily the energy has come from fossil fuels.
Data center energy usage is complicated - energy usage goes up and down with demand, so the amount they use is not constant.
Gas turbines solve this problem as well with what's called "dispatchable capacity" - power that can be turned on and off at will to match demand fluctuations.
Why Don't Data Centers Use Solar Panels?
Solar and wind power have major limitations for data center operations.
The fundamental problem is scaling. Solar panels and wind only generate power under the right conditions.
Data centers can't adjust their computing workloads based on weather patterns.
When a user turns to ChatGPT at midnight for a new pizza recipe, the servers need power immediately - not when the sun rises six hours later.
Battery storage is getting to the point where it could be a future solution, but current technology isn't economically viable for data centers, yet.
Natural gas turbines solve these problems by providing reliable, on-demand power that can be adjusted in real-time based on actual computing load.
This is why hyperscalers are signing contracts for gas turbines despite public commitments to renewable energy.
When uptime and reliability determine whether customers stay or leave, proven technology wins.
What Are Most Data Centers Powered By Today?
The energy mix varies by data center type and location, but a clear shift is underway.
Traditional data centers still draw most power from the electrical grid, accepting whatever energy mix their utility provides.
This typically includes natural gas, coal, nuclear, and some renewables depending on the region.
But new AI data centers - the facilities being built right now to support the AI boom - are increasingly incorporating on-site power generation.
Natural gas turbines are becoming the dominant choice for this on-site generation because they offer the fastest deployment timeline with proven reliability.
Two approaches are emerging:
On-site turbines: Individual gas turbines installed directly at the data center facility, generating power within minutes of demand spikes.
Energy islands: Self-contained power systems built near (but not directly adjacent to) data center clusters, serving multiple facilities with scalable capacity.
Both approaches bypass grid limitations while providing the dispatchable capacity AI data centers require.
The Innovation Shift: Where the Money Is Moving
Multi-year investments by data center operators into the gas turbine industry are boosting producers of industrial turbines and their supply chains.
This creates investment opportunities as companies building data centers sign long-term contracts with turbine manufacturers.
Deal backlogs are also growing for many companies creating the infrastructure and solutions for data centers with dispatchable energy gas turbines.
Our analysts believe this trend will continue until a better, more practical power solution emerges - which appears to be years away at this point.
That means years of revenue increases and share prices rising for the companies in this space providing the solutions to the AI power crisis.
The opportunity exists in two places for investors today:
Direct Turbine Manufacturers
If gas turbines are what data centers need, then the companies that manufacture them will benefit from this shift in the industry.
GE Vernova (GEV) leads this space with over 60 years of gas turbine experience dating back to the 1940s. The company can install turbines within days and begin generating power within 5 minutes.
GEV has partnerships with major natural gas suppliers like Chevron (the 7th largest natural gas producer globally by revenue) meaning they can also get the supply they need for these machines.
These relationships allow GEV to provide complete on-site power solutions to data center operators.
The company's shares are up more than 100% over the last 12 months as of October 23, 2025, and nearly 400% since spinning off from GE in April 2024.
Recent volatility stems from supply chain disruptions caused by tariffs and geopolitical conflicts, not from weakening demand.
Contract backlogs continue growing as hyperscalers commit to long-term power solutions.
And GE Vernova is a leader in this space.
Energy Island Builders
Beyond individual turbines, some data centers are implementing "energy islands" - self-contained power systems that serve specific sites or clusters without drawing from the larger grid.
Baker Hughes (BKR) specializes in these energy island solutions through its NovaLT product line.
The company has sold and tested these systems since 2013, with its first commercial data center turbine deployment in 2019.
Baker Hughes designs complete packages with turbines, generators, and systems tuned specifically for data center requirements.
The NovaLT runs on natural gas as well as hydrogen blends, providing some future-proofing as energy sources potentially evolve.
The stock has delivered steadier growth than more volatile turbine manufacturers - up 31% over the last 12 months as of October 23, 2025 and 226% over the last 5 years.
Supply Chain Opportunities
The turbine manufacturers represent direct exposure to this shift, but opportunities exist further down the supply chain as well.
Case and point: Gas turbines require replacement parts when they break down.
Plus, as technology advances, turbines need upgrades to handle increasing power demands and heat generation from next-generation AI computing.
The materials used to build these turbines become more critical as data centers push equipment to higher performance levels.
Companies producing advanced materials for turbine components may benefit as this market expands, particularly if they have diversified revenue streams that provide downside protection.
Understanding the Market Dynamics
The gas turbine opportunity for data centers exists because of a specific set of circumstances:
Immediate need: Data centers need power now, not in 5-10 years.
Grid limitations: Connecting to existing electrical infrastructure takes too long.
Proven technology: Gas turbines have decades of operational history and clear regulations.
Scalability: Production capacity can expand to meet demand without technological breakthroughs.
Reliability requirements: Uptime determines profitability for hyperscalers, making proven solutions more valuable than experimental ones.
This is a bridging technology. Nuclear or renewable energy sources might eventually provide better solutions one day
But those solutions aren't ready for deployment at scale today and gas turbines are.
The investment opportunity exists in the gap between current power needs and future technological solutions.
Companies positioned to supply power solutions right now benefit from hyperscaler spending that can't wait for future innovations.
What Could Go Wrong
Several risks could impact the natural gas turbine opportunity for data centers that investors will want to take note of:
Regulatory changes: Current lack of regulation around on-site power generation could change. New rules might slow deployments or increase costs.
Natural gas prices: Turbine economics depend on fuel costs. If natural gas prices spike, operating expenses increase for data center operators.
Alternative technologies: If nuclear power, renewable energy with storage, or other solutions become viable faster than expected, demand for gas turbines could decline.
Supply chain constraints: The supply chains supporting turbine manufacturing aren't currently strained, but dramatic demand increases could create bottlenecks.
Energy demand growth: If AI adoption slows or efficiency improvements reduce power consumption per server, the urgency driving this shift could diminish.
Investors need to understand this is a multi-year opportunity, not a multi-decade one.
What This Means for Investors
The data center power crisis creates a specific type of investment opportunity that innovation is aiming to solve.
Companies providing immediate power solutions benefit from hyperscaler spending that can't be delayed.
These aren't speculative bets on future technology - they're investments in proven solutions addressing urgent infrastructure needs.
The opportunity exists across several verticals:
- Turbine manufacturers signing multi-year contracts.
- Energy island builders providing scalable power systems.
- Material suppliers benefiting from equipment upgrades and replacement cycles.
Each layer offers different risk-reward profiles. Direct turbine exposure provides the most immediate impact but also the most volatility.
Supply chain plays offer more diversification but less direct exposure to the core trend.
And the timeline matters here - this is a bridge technology filling a gap while longer-term solutions develop. The investment window likely spans several years, not decades.
Investors will want to pay attention to contract announcements from data center operators because that will signal where money is flowing.
Backlog growth at turbine manufacturers indicates sustained demand, and policy changes around on-site power generation could accelerate or slow deployments.
Where Smart Money Is Moving
Every cloud backup, every AI query, every streaming service runs through a data center somewhere.
Natural gas turbines for data centers represent one Innovation Shift creating investment opportunities right now. But it's not the only one.
The energy crunch is one squeeze on your money. The dollar is the other. Join our CEO Jaspreet Singh on September 29th for a free workshop on how to profit from a dollar that is losing its value.
FAQ: Natural Gas Turbines and Data Center Power
What are most data centers powered by?
Most traditional data centers draw power from the electrical grid, receiving a mix of energy sources depending on their region - typically natural gas, coal, nuclear, and some renewables.
However, new AI data centers are increasingly using on-site natural gas turbines to bypass grid limitations.
These turbines provide reliable, dispatchable power that can be installed within months instead of waiting years for grid connections.
Where do data centers get their power?
Data centers get power from two primary sources: electrical grid connections and on-site generation.
Grid connections route electricity from centralized power plants through transmission lines. On-site generation uses natural gas turbines installed directly at or near the facility.
AI data centers are shifting toward on-site turbines because connecting to the grid takes years, while turbines can provide power within months.
Why is the U.S. building so many data centers?
The AI boom is driving unprecedented data center construction. Every AI application, cloud computing service, and digital platform requires physical servers in data centers.
Companies like Amazon, Microsoft, and Google are racing to build capacity to support AI services that generate billions in revenue.
Amazon Web Services alone generated $30 billion in Q2 2025. The U.S. is a hub for this buildout because hyperscalers want facilities near major population centers to reduce latency.
Why don't data centers use solar panels?
Solar panels can't provide the reliable, 24/7 power that data centers require. Solar only generates electricity when the sun shines, creating supply and demand problems.
Data centers need power at all hours regardless of weather conditions - you can't schedule AI processing for only sunny days.
Battery storage large enough to power a facility consuming as much electricity as 80,000 homes isn't economically viable with current technology.
Natural gas turbines provide on-demand power that adjusts to actual computing load in real-time.
Are data centers using natural gas?
Yes, and more data centers are turning to this source of energy.
Data center operators are signing long-term contracts with gas turbine manufacturers for on-site power generation.
The gas turbine market for power generation is growing from $16.7 billion in 2023 to a projected $30.4 billion by 2034, with much of this growth driven by data center demand.
Natural gas turbines solve the immediate power problem while nuclear and renewable solutions remain years away from viability at scale.
What type of energy do data centers use?
AI data centers need "dispatchable capacity" - power that can be turned on and off at will to match demand fluctuations.
Natural gas turbines provide this flexibility, generating continuous power that adjusts in real-time based on computing load.
Traditional data centers use whatever mix the local grid provides, but new AI facilities are implementing on-site gas turbines or energy islands to bypass grid limitations and ensure 24/7 reliability.
How much power does a data center use?
A single AI data center can consume enough electricity to power 80,000 homes.
The U.S. Department of Energy projects data center energy consumption will double or potentially triple by 2028.
AI computing requires exponentially more power than traditional cloud computing because the servers generate more heat and require more cooling, creating a compound effect on total energy consumption.
Which companies benefit from data center power demand?
Companies across the natural gas turbine value chain benefit from growing data center power needs.
Direct turbine manufacturers like GE Vernova are signing multi-year contracts with hyperscalers.
Energy island specialists like Baker Hughes provide complete power systems for data center clusters.
Advanced materials companies supplying turbine components benefit from equipment upgrades and replacement cycles as data centers push technology to higher performance levels.
The AI Power Crisis & Its Data Center Energy Solution
AI data centers face a power crisis that the electrical grid is having trouble solving today
Natural gas turbines provide immediate relief, offering reliable power within months instead of years.
Hyperscalers need this power - and are committing billions to this proven technology because their businesses depend on 24/7 uptime.
The gas turbine market is growing from $16.7 billion in 2023 to $30.4 billion by 2034.
Companies positioned to supply power solutions today - turbine manufacturers, energy island builders, and supply chain specialists - may benefit as data center operators sign long-term contracts worth billions.
One day, nuclear or renewable energy sources may be the main power suppliers to data centers. But today, its gas dispatchable energy gas turbines.
Our analyst believes this opportunity will last for the next few years as other energy sources scale.
But for investors who identify this Innovation Shift now, the opportunity to profit from AI's power crisis exists in the near term.
