Aschenbrenner Fund Buys Nebius After Nvidia Deal

Leopold Aschenbrenner’s fund bought Nebius after Nvidia pledged $2 billion to the company.

AI infrastructure has become one of the hottest corners of the market, and Nebius Group is now sitting right in the middle of it. On May 28, the stock was trading around $226 after a huge run, and the latest disclosure showed that Situational Awareness Fund had taken a 5.6% stake in the company.

That move matters because the fund is run by Leopold Aschenbrenner, the former OpenAI researcher who became widely known for his long essay on AGI and the years ahead. Now he is putting money behind the infrastructure layer that powers AI models, and the market is paying attention.

Why Nebius is getting so much attention

Nebius is a neocloud provider, which means it is built for one job: renting out high-performance GPU capacity for AI workloads. That is different from the broad cloud model used by Amazon Web Services, Microsoft Azure, and Google Cloud, which serve everything from databases to business apps.

The AI boom created a very specific bottleneck. Companies want compute for training, inference, and deployment, but the biggest cloud providers cannot always supply enough GPU capacity fast enough. Nebius built its business around that shortage, and investors are treating that specialization as an asset rather than a niche.

• Q1 revenue rose 684% year over year.
• EBITDA beat expectations by 43%.
• Adjusted EPS beat expectations by 50%.
• The stock jumped 24% in the days after that earnings report.

The market also likes the fact that Nebius is not just talking about demand, it is already signing large contracts and expanding aggressively. The company has said it wants to deploy more than 5 gigawatts of AI computing capacity by the end of 2030, including multiple gigawatt-scale AI factories in the United States.

That is an expensive plan, but it also explains why investors keep circling the stock. In AI infrastructure, scale is the moat. If a company can secure power, chips, and data center capacity early, it can sell compute into a market that still looks supply-constrained.

Nvidia’s backing changes the story

Nvidia is not making a casual bet here. The company plans to invest $2 billion in Nebius by 2030, and the two companies are working together on an inference stack for developers and businesses. Nebius also plans to use Nvidia’s next-generation hardware, including Rubin GPUs, Vera CPUs, and BlueField storage systems.

That matters because Nvidia is helping shape the operating layer, not just selling chips. The collaboration also includes monitoring tools for GPU health and software optimization recommendations, which makes the relationship deeper than a simple supplier contract.

“We are entering the age of AI factories,” Jensen Huang said in a 2024 keynote at CES.

That quote has aged well. Nebius is effectively trying to build one of those factories at scale, and Nvidia’s support gives the plan more credibility than a typical startup pitch ever could.

The stock has already reflected that optimism. Shares are up 103% since the collaboration was announced in March, and over the last 12 months the gain is close to sixfold. That kind of move usually forces investors to ask the harder question: what is already priced in?

The numbers behind the rally

There is a real business story underneath the excitement. Nebius generated less than $900 million over the last four quarters, yet Wall Street is now modeling $3.5 billion in revenue for 2026 and $11 billion for 2027. That is a steep ramp, and it shows just how much growth investors expect from AI infrastructure.

At the same time, valuation is no longer cheap in the casual sense. Using the article’s cited figures, Nebius trades at forward price-to-sales ratios of 16.6 for 2026 and 5.3 for 2027. Those numbers are high, but they are also tied to a company that is still in the middle of a buildout, not a mature utility business.

• Market cap is roughly $58 billion in the article’s figures.
• Current price was about $226 per share.
• 52-week range ran from $34.72 to $233.73.
• Average volume was 17.3 million shares.

There are risks. Power is one of the biggest. Nebius recently signed a $2.6 billion deal with Bloom Energy for 250 megawatts of power capacity and 328 megawatts of installed capacity over 10 years, but that is still only a slice of what the company will need.

Competition is the other risk. Well-funded hyperscalers can build AI-specific infrastructure too, and they have balance sheets that make it hard for smaller players to sleep comfortably. Nebius has an early lead, but it will need to keep turning capital into usable capacity faster than rivals can copy the model.

What investors should watch next

The Aschenbrenner purchase says as much about the AI market as it does about Nebius. Investors who once focused on model labs and chipmakers are now moving down the stack to the companies that provide power, GPUs, and data center capacity. That is where the bottleneck is, and bottlenecks tend to get priced aggressively.

If Nebius keeps showing strong revenue growth while improving profitability, the stock can justify a lot of optimism. If margins start to compress before the company scales enough, this rally could cool fast. The next few quarters should tell us whether Nebius is building a durable AI utility or just riding a very expensive wave of demand.

For now, the key question is simple: can Nebius keep expanding capacity fast enough to stay ahead of both customer demand and larger cloud rivals?

Related reading: Why Nebius stock moved sharply higher today and why Nebius has soared over the past year.