Hyperscalers are doing everything they can to secure energy, putting this utilities firm in a position to benefit heavily
The race for reliable access to energy has raged as hyperscalers compete against each other in the build-out of data centers and other AI-related infrastructure.
As a result, tech firms are looking into alternative sources of power like nuclear to keep up with AI’s gargantuan energy demands.
Microsoft and Meta have entered into deals with Constellation Energy to restart nuclear power plants located in Three Mile Island and Clinton, Illinois. However, those tech firms aren’t the only ones making such moves.
Earlier this week, Alphabet announced that it was partnering with NextEra Energy (NEE) to restart a nuclear power plant in Iowa that was shuttered due to storm damage in 2020.
Deals like this one exemplifies AI firms doing whatever they can to keep up with the AI boom’s massive energy demand.
Investor Essentials Daily:
Friday News-based Update
Powered by Valens Research
Hyperscalers are doing everything they can to build data centers capable of performing artificial intelligence (“AI”) workloads. As a result, this build-out is expected to cost trillions in capital expenditure.
As AI firms race to develop advanced AI models and release them to the public, they will need to eliminate another bottleneck: energy.
In 2022, Federal Energy Regulatory Commission (“FERC”) forecast that peak energy demand in 2029 would rise 23 gigawatts (“GW”) from 2022 levels to 840 GW. A year later, FERC revised its estimate, projecting peak demand of 859 GW by 2029.
Due to AI workloads consuming lots of power, it’s expected that data centers are expected to consume as much as 12% of the electricity generated by the U.S. grid by 2030.
As a result, the FERC revised its estimates, forecasting that peak energy demand will surge to 947 GW by 2029.
With AI consuming so much energy, many industry experts are pointing to nuclear energy as a potential solution to growing energy demand. And this opportunity hasn’t been lost on hyperscalers who are racing to build—or restart—nuclear power plants.
This scenario has played out with Constellation Energy (CEG), which partnered with Microsoft (MSFT) and Meta (META) to restart nuclear power plants in Three Mile Island and Clinton, Illinois, respectively.
However, this energy provider isn’t the only one benefiting from this tailwind, and Microsoft and Meta aren’t the only hyperscalers making such moves.
Earlier this week, it was announced that Alphabet (GOOGL) is partnering with NextEra Energy (NEE) to restart the Duane Arnold Energy Center, a 615-megawatt nuclear facility in Iowa that was shuttered after sustaining damage in 2020.
NextEra is a leading renewable energy company in the U.S.. It’s also the owner of Florida Power & Light Company, which provides power to roughly 12 million people in the state.
Under the terms of the agreement, Alphabet signed a 25-year agreement to acquire energy from the shuttered Iowa plant once it restarts operations. After being inoperable for five years, it’s estimated that it will cost $1.6 billion to restart.
The Duane Arnold nuclear facility is expected to start providing power by 2029, according to NextEra who will lead the redevelopment of the plant. This has also resulted in Alphabet securing roughly 3 GW of power from NextEra.
In addition, Alphabet announced that it agreed to explore opportunities with NextEra in the deployment of new nuclear power generation capacity across the U.S.
In the aftermath of the deal, the energy firm revealed its plans to further develop its power generation capacity in Iowa.
This recent partnership exemplifies AI firms doing whatever they can to keep up with the AI boom’s massive energy demand. As a result, firms like NextEra will be uniquely positioned to capitalize.
As the AI arms race continues, investors should keep an eye out on energy utilities like NextEra because they provide the electric power desperately needed by hyperscalers.
Best regards,
Joel Litman & Rob Spivey
Chief Investment Officer &
Director of Research
at Valens Research