So, let's get straight to it, because today we saw something that almost nobody expected. Meta's stock surged more than 10% in a single day. The reason? Reports that the company is building its own cloud business to sell excess computing capacity.
So far, so good. Great news for Meta.
The strange part? That very same news sent Nvidia, CoreWeave, ASML, and dozens of other AI related stocks sharply lower. In other words, one piece of positive news for a single company ended up shaking the entire artificial intelligence sector.
Let's start from the beginning. Meta is spending enormous amounts of money on artificial intelligence. How much? Back in April, management told investors it plans to spend as much as $145 billion in capital expenditures this year. Data centers, graphics processors, infrastructure, and everything in between.
Now comes the new move. Instead of letting all that computing power sit idle whenever it is not being used internally, Meta wants to rent it out to other companies. Internally, the project is reportedly called "Meta Compute."
What does that actually mean? Simply put, Meta wants to become a cloud provider by leasing computing capacity to outside customers.
The company is reportedly considering two approaches. The first is offering access to AI models running on its infrastructure, similar to Amazon's Bedrock platform. The second is selling raw computing capacity, much like CoreWeave does today.
Meta would not be the first company to monetize unused infrastructure. Other large technology companies have already shown that excess computing capacity can become a valuable source of revenue through large commercial contracts.
This is where things become more interesting. Why were investors so excited?
For months, many investors had been worried about Meta's massive spending. They watched billions of dollars flow into AI infrastructure and kept asking the obvious question: when will all of this investment start generating meaningful revenue?
This announcement may offer part of that answer. If Meta can sell computing power that would otherwise remain unused, it can begin generating revenue from assets that until now represented only a significant expense.
That was also the view expressed by UBS analyst Stephen Ju. According to him, selling excess computing capacity could reduce earnings risk because it creates revenue immediately instead of waiting years for Meta's AI products and business agents to mature.
There is, however, one important caveat. The same analyst pointed out that some investors may question why this cloud business was never highlighted by Mark Zuckerberg as one of the company's long term strategic priorities.
Now let's look at the other side of the story.
While Meta was soaring, much of the AI sector turned sharply lower.
Why would positive news create such a negative reaction?
The answer lies in one word: oversupply.
To understand this, think about why these companies rallied so aggressively over the past year. Demand for AI computing power dramatically exceeded available supply. Companies simply could not build infrastructure fast enough, allowing providers to charge premium prices.
Meta's announcement raised a new concern. If another major player starts offering large amounts of computing capacity, perhaps supply is finally beginning to catch up with demand. And if excess capacity enters the market, pricing power could weaken, putting pressure on profit margins across the industry.
The market reacted quickly. CoreWeave dropped more than 10%, Nebius nearly 12%, Nvidia around 2%, AMD close to 3%, and memory manufacturers such as Micron fell roughly 7%. Equipment makers including ASML, KLA, Lam Research, and Applied Materials also declined, along with networking companies such as Corning.
The reaction was especially striking because many of these stocks had enjoyed enormous gains in recent months. AMD had climbed more than 185%, while Micron had surged roughly 241%. That shows just how quickly investor sentiment can shift after a single unexpected development.
It is also worth noting that the selloff was highly targeted. The Nasdaq fell only about 0.33%, the S&P 500 was almost unchanged, and the Dow Jones actually finished higher. In other words, this was not a market wide panic. It was a sharp reassessment of the artificial intelligence sector.