The news media and investors are rapidly realizing that the growth of artificial intelligence (AI) will require massive increases in available electrical power. And investors have a significant opportunity to profit from AI’s growing need for power.

A recent Forbes article titled, “AI Is Pushing The World Toward An Energy Crisis,” includes some interesting facts.

• GPT-4 required over 50 gigawatt hours to train and operate. This was 50 times the power it took to train GPT-3.
• Data center power consumption is forecast to grow from 460 terawatt-hours in 2022 to 1,000 terawatt-hours in 2026. That level equates to 6% of all power used in the U.S.
• Data center energy demands are expected to double again by 2030.
• Northern Virginia, with 51 data centers and 51 million square feet of dedicated space, has the largest concentration of these centers. Those data centers consume electricity equivalent to 800,000 homes.
• A rack of traditional computer servers runs on seven kilowatts. A rack of AI servers uses 30 to 100 kilowatts.

The article notes that nuclear and natural gas are the only two possible sources of added power to support AI.

Proponents of nuclear power look forward to the development of small modular reactors (SMRs) to provide on-site or local power for data center concentrations. Unfortunately, the development of SMRs is in the early stages, and it will be in the 2030s before they are ready for operational deployment.

Which leaves us with natural gas. This energy source is already the most significant fuel for electricity in the U.S. If public utilities provide the added power needed by data centers, they will need more natural gas. There is also the potential for developing SMR-type natural gas power plants.

I like investing in natural gas by investing in gas gathering and transport infrastructure. These are the gathering, processing, pipeline, and terminal companies. Liquified natural gas (LNG) processing is also an essential part of the story. Here are a couple of investment ideas:

The InfraCap MLP ETF (AMZA) gives investment exposure to the largest MLPs, which control a large portion of the energy midstream infrastructure. AMZA yields 7.25% and is growing the monthly dividend by 8% to 9% annually.

Cheniere Energy (LNG) is the world’s largest producer of LNG. The company will benefit from the global growth in the use of LNG as well as natural gas needed to power data centers.

For something fun, New Fortress Energy (NFE) recently announced the launch of Klondike Digital Infrastructure, a power and data center development business. According to the press release, Klondike is revolutionizing the industry by utilizing behind-the-meter, on-site power. Also in the press release was this:

Klondike is currently developing a geographically diverse portfolio of data center sites to satisfy the requirements of hyperscale users. The company has more than 1,000 acres of developable land across sites in Brazil, Ireland, and the United States. These locations have large existing power plants or permits in process to build several gigawatts of power, connectivity to fiber networks, access to transmission and water.

NFE is a portfolio stock in my Monthly Dividend Multiplier service. I am on the hunt for other companies with similar projects.t with some dividend growth, and you will have an excellent core income portfolio holding.