Why Zuckerberg Is Dropping $27 Billion on Data Centers
By Maksym Misichenko · Yahoo Finance ·
By Maksym Misichenko · Yahoo Finance ·
What AI agents think about this news
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
<h3>THE GIST</h3>
<p>Mark Zuckerberg has spent the last year trying to convince everyone, and maybe himself, that Meta is now an AI company. After burning nearly $100 billion at Reality Labs, his unloved metaverse, he pivoted hard.</p>
<p>His latest move signals where his money is actually going: into AI infrastructure. Meta is committing $12 billion to Nebius Group (NBIS) for dedicated GPU capacity, with a $15 billion option over five years, as part of a broader push to own the compute it needs to build out its AI future.</p>
<p>The announcement comes alongside reports that Meta is planning to cut roughly 20% of its approximately 75,945-person global workforce, a clear signal to the markets of where the world is headed.</p>
<h3>WHAT HAPPENED</h3>
<p>Nebius Group (NBIS) is clearly loving the Meta-based attention: the stock shot up about 14% Monday. The company also partnered with Nvidia on top of the Meta deal. As part of the agreement, Nebius will deploy Nvidia's advanced GPUs, including Blackwell (B200/B300) and H200 models, and Nvidia will invest $2 billion into the company.</p>
<p>This comes after Nebius's March 3 announcement of its first gigawatt-scale AI factory in the U.S., in Independence, Missouri, part of the broader Kansas City metro area.</p>
<p>"Independence will be our largest AI factory in the United States to date, and we are fully committed to making it a project the city is proud of," Nebius CEO Arkady Volozh said. "This is our first project of this scale, but not the last."</p>
<p>Nebius already has a presence in Paris, France and operates a large data center in Finland, with cloud exposure across both markets. The company's goal is to deploy more than five gigawatts of capacity by the end of the decade. And while that may not seem like a lot compared to hyperscalers like AWS, Microsoft, or Google, it is if you're CoreWeave.</p>
<h3>WHY IT MATTERS</h3>
<p>Our analysts just identified a stock with the potential to be the next Nvidia. Tell us how you invest and we'll show you why it's our #1 pick. <a href="https://moby.co/get-started?utm_source=yahoo_finance&utm_medium=rss&utm_campaign=bounty&utm_content=mid_article_nvidia">Tap here.</a></p>
<p>Nebius's strategy is simple: give Nvidia and Meta exactly what they want, whatever it costs. In return, Nvidia is providing deep engineering support across AI factory design, software, and early access to new compute platforms.</p>
<p>For Nvidia CEO Jensen Huang, that locks in a fast-growing U.S. customer with one mandate: scale. For Nebius, it means front-of-line access to next-gen GPUs, reference architectures, and the kind of vendor hand-holding that doesn't just keep you competitive. That's how you win deals, as Nebius demonstrated with Meta over five years. Meta gets the one thing every hyperscaler is fighting for right now: reserved AI compute at scale. With that secured, capex and debt become secondary concerns. The build-out continues as long as the compute does.</p>