In a research note released late Thursday, Cantor Fitzgerald says Core Scientific (CORZ) could fetch over $30 per share in a potential acquisition by cloud compute giant CoreWeave, citing both the long-term cash flows from its AI contracts and the replacement value of its data centers.

That would be a near doubling from the current level just above $16.

The note came hours after The Wall Street Journal reported that CoreWeave, a cloud AI compute firm, is once again in advanced talks to acquire Core Scientific, following a failed $5.75 per share offer in 2024.

CORZ shares jumped 33% to close over $16 Thursday, but Cantor believes that it still undervalues the company by at least 50%.

At the heart of the bull case is a 12-year, $3.5 billion infrastructure lease Core Scientific signed with CoreWeave in 2024 to provide 200 megawatts of AI capacity.

Cantor values the lease stream at $24/share, using a conservative 15x profit multiple typical for traditional data center REITs. Add another $11.70/share for the replacement value of CORZ’s 570MW of power infrastructure, and the upside case becomes clear.

The BTC — AI Pivot

But it’s not just Cantor arguing that the compute power used for crunching numbers to mine BTC might be more efficiently used for AI.

Rittenhouse Research, a new fintech and AI-focused firm, released a report in May arguing that the most successful crypto companies aren’t doubling down on bitcoin. Instead, they’re pivoting to become AI infrastructure providers.

When Galaxy Digital bought the Helios data center in late 2022, it seemed like a rescue of a struggling miner, yet it turned out to be a strategic AI asset as demand for data center space surged with the rise of ChatGPT and LLMs, Rittenhouse pointed out.

“The infrastructure used to mine digital gold is better used to process AI algorithms,” Rittenhouse wrote at the time.

At the core of the argument is the belief that AI generates stable, long-term cash flows, unlike BTC mining, which is subject to sharp revenue drops every four years due to halvings and is heavily dependent on bitcoin’s volatile price cycles.

The future profitability of BTC mining, Rittenhouse noted, is also dependent on mining firms being able to design chips that are significantly more efficient each cycle to account for the halvening, an increasingly difficult task as gains from silicon shrinkage begin to plateau.

But Not Every Pivot Away from BTC is Successful

While Cantor, and the market broadly, is looking fondly on Core Scientific’s possible pivot, not all pivots away from BTC mining have gone this well.

As CoinDesk recently reported, Bit Digital is dumping its bitcoin rigs to go all-in on Ethereum staking, and the market pushed down its stock by 15% during the Thursday trading session in New York.

Canaan, once hoping to diversify into AI hardware, has now shuttered its chip unit entirely after failing to gain traction. Its stock is down nearly 75% in the last six months, and closed at 63 cents on Thursday.

But Core Scientific might have found the middle path, leveraging its mining-built footprint to tap into a $100 billion-plus AI infrastructure boom.

If Cantor’s thesis proves right, CoreWeave’s second offer for CORZ could look very different from the one they made last year, and it could mark a new blueprint for the rest of the sector.

Neither CoreWeave nor Core Scientific has publicly commented on the matter.