AI Investing Genius Leopold Aschenbrenner Owns 20% of This AI Stock. Should You Buy It Too?
SharonAI Holdings (SHAZ -7.53%) got on more people’s radars when artificial-intelligence (AI) investing genius Leopold Aschenbrenner revealed that his hedge fund owns 20% of its shares.
Aschenbrenner’s Situational Awareness hedge fund returned more than 2,000% in 2025 and has more than doubled so far this year. That’s why many growth investors pay attention when he makes a move.
A small stake in SharonAI would have been enough to draw some attention, but a 20% position? SharonAI has certainly done its part for Aschenbrenner so far. It has more than doubled year to date, and further gains may be around the corner.
Image source: Getty Images.
A power shortage play
SharonAI is a neocloud provider that leases out data center space powered by its own AI chips. This setup is very similar to CoreWeave (CRWV -11.58%) and enables the company to scale without building data centers from scratch.

Today’s Change
(-7.53%) $-5.29
Current Price
$64.89
Key Data Points
Market Cap
Day’s Range
$62.21 – $71.69
52wk Range
$1.85 – $178.00
Volume
1.7M
Avg Vol
1.2M
This model lets SharonAI scale with less overhead, but not owning land can be costly in the long run. However, megawatts are extremely valuable right now, especially for neocloud providers that have their own chips. That value proposition helped SharonAI land Canva, GMI Computing, and an unnamed U.S. company as new customers in the first quarter.
The company is still small but growing quickly. Q1 results revealed that it hoped to increase data center capacity from 50 megawatts to 100 megawatts by early 2027. Meanwhile, a July press release revealed that it has 116 megawatts of contracted capacity, marking a substantial increase in a short amount of time.
Megawatt capacity is expected to be a major AI theme in the years ahead. As companies create more AI-enabled products and attract more customers, they will need more megawatts to keep up with customer demand. SharonAI is well positioned to address this bottleneck.
Comparing SharonAI to CoreWeave
CoreWeave is the main comparable to SharonAI, so it’s good to assess CoreWeave’s megawatt pipeline, market cap, and long-term goals. The company has more than one gigawatt of active power, roughly nine times more than SharonAI. CoreWeave’s $45 billion market cap is also 18 times bigger than SharonAI’s $2.5 billion market cap.
If SharonAI traded at the same valuation at CoreWeave based on contracted megawatts, the stock would have to double from its recent price of about $72. However, CoreWeave has forecast more than eight gigawatts of active power by 2030. It’s an aggressive goal, but the fact that CoreWeave is pushing toward it explains the premium.
SharonAI has demonstrated that it can accumulate megawatts much faster than initial projections. Aschenbrenner seems to be betting heavily on that scenario.
Not owning land makes it riskier in the long run
Investors who like CoreWeave may also like this growth stock, but there is a significant risk in not owning the land. Landlords can charge much higher rates when it’s time to renew leases. That can pressure margins, which remain a concern. SharonAI is still unprofitable at this time.
Iren (IREN -8.65%) and Nebius (NBIS -15.02%) strive to own their land, data centers, AI chips, and power. Co-location companies like Cipher Mining (CIFR -10.34%) and TeraWulf (WULF -7.93%) own their land, data centers, and power, but not AI chips.
The co-location model involves having a customer bring their own chips, while SharonAI has the chips but not the land. Aschenbrenner is well-diversified across neocloud plays, but Nebius is by far his largest position. It makes up 40% of his hedge fund’s portfolio, while a little more than 6% of the fund’s capital is in SharonAI.
Notably, Aschenbrenner has a larger percentage of SharonAI shares than Iren and Applied Digital (APLD -9.03%). This approach can work if SharonAI expands its contracted megawatts faster than anticipated and improves its margins, but it is a high-risk stock, which has become the norm in the artificial intelligence industry.