“No power, no AI. Or to put it more clearly: no electricity, no progress.”
That’s what the CEO of Siemens Energy told shareholders at a recent meeting.
You see, everyone is excited about the potential of artificial intelligence (AI). But they’re forgetting one important thing.
The rapid increase in demand for AI is running into a harsh reality… Our power grids can’t handle the massive increase in electricity needed to power the chips that make AI work.
AI is an energy hog.
Talking to ChatGPT uses about 10 times more energy than running a search on Google. And generating one image uses enough power to drain a smartphone battery by 25%.
That’s becoming a problem.
Data centers, which power the internet, are already struggling to get enough electricity to run all their computers.
So this will only get worse as more people start using AI and companies develop more complicated AI programs that demand more power.
Here at Intelligent Income Daily, we’re focused on finding the safest income investments on the market. When we spot major trends in an industry, we pay attention… and look for ways to profit.
Today, I’ll explain why data centers can’t get enough power and why the problem won’t go away anytime soon. I’ll also show you one way to profit from this situation while collecting a reliable yield.
The Internet’s Brain Needs a Lot of Power
If the internet had a brain, it would be in Virginia.
That’s because the state is home to the largest cluster of data centers in the world – nearly 48 million square feet of computing power.
Nobody knows for sure how much of the world’s internet data flows through Virginia. But it’s a lot. Estimates range from 30% to 70%.
Here’s where the problem comes in…
Those computers in Virginia use a huge amount of power – more than 3,400 megawatts. That’s enough to power over 2.5 million homes.
And that’s causing problems for the local power grid.
Electricity produced at a power plant must flow through a network of power transmission lines before it reaches the homes and businesses where it’s used.
But each power line can only handle so much electricity. If it gets overloaded, the massive amount of heat created from the electricity flowing through can melt the wire.
Data centers in Virginia are using so much power that they need more transmission lines. Unfortunately, those take years to build and come with a hefty price tag.
So it’s no surprise that over the past year, electricity rates have soared in parts of the country with lots of data centers. Look at the chart below to see how much prices have increased from 2022 to 2023.
Tech companies used to build data centers in cities with major internet hubs to get the fastest connections.
But now, it’s getting harder to find enough power to keep up with the demand from AI.
A recent analysis in the journal Joule estimates that over the next few years, AI server units could add up to 15,000 megawatts of electricity demand.
That’s more power than what Bitcoin miners currently use.
And it doesn’t even account for the energy needed to cool the data centers. Data centers with AI servers need 10 times more cooling per square foot compared to those with traditional CPU servers.
All combined, that’s a lot of energy demand.
How to Profit From This Growing Demand
According to Boston Consulting Group, electricity demand from data centers could rise to 390 terawatt-hours by 2030.
That’s triple the amount data centers used in 2022. And enough electricity to power 40 million homes.
Tech companies will have to figure out how to get all that electricity without overloading the power grid.
And one company helping data centers get the power they need is AES Corporation (AES).
AES is one of the largest renewable energy developers in the country. It has signed contracts with companies like Alphabet and Microsoft to build wind and solar farms to provide clean electricity to data centers.
AES expects to install 12.3 gigawatts of renewable power over the next three years. That’s enough to power 9 million homes.
According to CEO Andres Gluski, AES has figured out how to use a mix of batteries, solar, wind, and hydroelectric power to make sure data centers have stable, clean power all the time. That’s expertise that tech companies are willing to pay up for, allowing AES to make better profits.
And best of all for us, AES has increased its dividend 13 years in a row and plans to keep increasing its payout. Right now, it yields 4.6% and trades at less than 8x earnings.
That’s 30% below its historical average of 13x earnings. So it’s a good time to invest at a discount.
The demand for electricity will only increase as AI and its applications grow. But you don’t need to bet on a tech stock to profit from this growth.
By helping power the entire trend, you can earn reliable income for years to come.
Happy SWAN (sleep well at night) investing,
Brad Thomas
Editor, Intelligent Income Daily