A $9 billion AI infrastructure acquisition deal unexpectedly failed in a shareholder vote. U.S. Bitcoin mining company Core Scientific officially rejected a stock-only takeover offer from its competitor and partner, CoreWeave, on Thursday. This decision not only shocked the capital market but also reflected the huge differences and speculative mindset among investors regarding the valuation of computing power assets in the current AI boom.
The key driver behind this rejection was Sina Toussi, the founder of Two Seas Capital, Core Scientific's largest shareholder, who focuses on investing in post-bankruptcy companies. In an open letter to shareholders, he stated: "Since the transaction was announced in July, AI infrastructure investment has accelerated, and peer company valuations have repeatedly hit new highs. Why accept a bid of only $16.40 per share?" In his view, Core Scientific had the potential to grow independently into a "second CoreWeave."

From Mining Farms to AI Centers: Two Companies with a Common Origin but Different Paths
Core Scientific and CoreWeave share a common starting point — both were initially cryptocurrency mining companies, accumulating data center assets through high-density computing facilities. However, as market trends changed, CoreWeave quickly shifted toward the AI cloud computing sector and received support from major investors such as NVIDIA. Since its IPO, the company's market value has surged from $14 billion to $66 billion (a stock price of about $140), making it a popular target for capital markets betting on AI infrastructure.
In fact, even while pursuing the acquisition, CoreWeave had already signed a 12-year, $1 billion long-term contract with Core Scientific to rent its data centers for deploying AI computing services. The July-announced acquisition was seen as a natural evolution from cooperation to integration, but it ultimately fell through due to a gap in valuation expectations.
After Rejecting the Acquisition, Core Scientific's Market Value Rose to $6.6 Billion
After the shareholders rejected the acquisition, Core Scientific's stock price immediately jumped, and its current market value has reached $6.6 billion, far exceeding the value implied by the original offer. This indicates that the market is betting that, in the context of continued explosive demand for AI computing power, Core Scientific, with its existing data centers and power resources, could achieve higher valuations by building its own AI cloud services.
However, this gamble also carries risks. Analysts point out that investors frequently rejecting acquisition offers and expecting higher premiums is a typical sign of AI asset bubble formation — market sentiment is overheated, and valuations are detached from short-term fundamentals.
CoreWeave Turns to Acquire an Open Source Tool Company, Advancing to the Upper Application Layer
After being rejected, CoreWeave did not stop its expansion. On the same day, the company announced the acquisition of Marimo, a competitor of the open-source Jupyter Notebook, without disclosing the transaction amount. According to PitchBook data, Marimo had previously raised about $5 million in funding.
This move sends a clear signal: CoreWeave is moving from underlying computing power hosting to the AI application development ecosystem. Jupyter Notebook-type tools integrate code, text, and visualization, and are widely used in data science and AI model development. By integrating Marimo, CoreWeave is expected to build a full-stack service capability of "computing power + development environment + deployment," further solidifying its position in the AI developer ecosystem.
Beneath this acquisition controversy lies a critical turning point in the AI infrastructure sector, shifting from "seizing territory" to "refining operations." Whether through independent breakthroughs or integration and expansion, the ultimate outcome for computing power players will no longer be determined solely by the number of cabinets, but by who can truly connect the value chain from electricity, chips to developer experience. For now, this gamble has just begun.
