Hyperscalers' AI buildout will require massive amounts of energy. Two under-the-radar stocks will benefit

LivestreamMenuMake ItselectUSAINTLLivestreamSearch quotes, news & videosLivestreamWatchlistSIGN INCreate free accountMarketsBusinessInvestingTechPoliticsVideoWatchlistInvesting ClubPROLivestreamMenu (This is CNBC's "Power Insider" newsletter, your inside look at the investments, people and companies powering the global energy industry. Click here to subscribe.) Hyperfocused on hyperscaler hypergrowth. I tried to sum up investor thinking into just one sentence. The pace of spending on A.I. is so frenetic it makes the Energizer Bunny look lazy. That growth is coming from the "hyperscalers," just a fancy term for the big technology companies that are rapidly ramping their bets on artificial intelligence. This energy-related weekly intelligence piece is about artificial intelligence because - in this author's reasonably humble opinion - there isn't an A.I. story without energy. A.I. requires massive amounts of computational power - "compute" - and compute requires massive amounts of electricity. In other words - and say it with me - A.I. is power. Literally. As long as the A.I. spend story steams along, it seems logical that the investment in energy will steam with it. Now that we are coming out of another earnings cycle, three things remain clear: 2. Capital spending related to A.I. is a big part of that story "AI Hyperscalers capex continues to be revised higher. Following issuers' guidance at 1Q earnings season, estimates for 2026 capex are now $725bn, this has nearly doubled since mid-2025. Capex is rising faster than OCF [operating cash flow], driving funding needs. " The numbers are hard to fathom. BNP Paribas highlights that consensus estimates for capital spending were for 'just' $365bn one year ago, which means this year's capex estimate of $725 billion is nearly double last year's estimate. When was the last time you saw a major estimate nearly double in a year? Let's put that $725 billion in A.I. related spending growth into perspective. $725 billion is more than the total GDP output of some mid-sized European countries and about 1 ½ times bigger than the economy of Singapore. $725 billion is roughly the same size as the market cap of JPMorgan Chase. It's only about $125 billion less than the entire value of ExxonMobil and twice the value of Chevron. $725 billion is over 3 times more than the value of every NFL team — combined (check out CNBC's exclusive NFL valuation analysis here). You may see other numbers. Each firm has their own estimates. They are all, however, very bullish. For example, UBS sees nearly half a trillion being spent, and that is only on the power side. They write: "Overall, we see $511bn being spent by 2030 on generation capacity additions for a 3% [compounded annual growth rate], which does not include transmission or distribution build out." UBS believes that if this kind of spending continues, both natural gas and solar will continue to see "sold out order books." Evercore ISI is even more optimistic, seeing about $800 billion in spending, with most of that coming from Alphabet (GOOGL), Microsoft (MSFT), Meta (META), Amazon (AMZN) and Oracle (ORCL). $700 billion, $800 billion, etc etc. However much the final numbers end up being, they are colossal. The bottom line: You rarely know you are making history while living through it. My friends, this is history. We are making it in real time. It reminds me of when I was starting out in this business and the internet was just growing up. Companies will come and companies will go, but this investment cycle is real and it's spectacular. Remember though, like with most Wall Street history, there will be winners and there will be losers. Some big winners and some stocks that get crushed. Stay focused. Keep watching and reading CNBC. And enjoy the ride. TAKE ACTION → So how can you invest in and around this massive A.I. capex cycle? One way is of course to invest in the hyperscalers themselves. The super mega cap names you may already be invested in. Another is to look at companies that provide the power for all these A.I. dreams. One of those is Hut 8 (HUT). The Miami-based energy infrastructure company keeps making investors a ton of money. Last week Hut 8 signed a $9.8 billion dollar deal and the stock soared. We interviewed CEO Asher Genoot about the massive deal and you can watch it here. The analysts that cover Hut 8 have a $118.13 target on the stock. Another example is smaller cap Fluence Energy (FLNC). The energy storage and battery company posted a narrower loss and signed supply agreements with two big hyperscalers. That news sent the shorts running to cover and the stock soaring. Shares doubled in a week. But investors take note: the stock price is now above the current 12-month price target for Wall Street analysts. UBS likes companies that benefit in other ways from all this spending. Its analysts believe Eaton (ETN) and Brazil-based WEG (WEGE-BR) have 'tailwinds' from the expected power generatio

Hyperscalers' AI buildout will require massive amounts of energy. Two under-the-radar stocks will benefit

Key takeaways

More in business