AI Cloud Provider CoreWeave Acquires Core Scientific for $9 Billion to Secure Data Center Capacity

CoreWeave, the specialized artificial intelligence cloud provider, has reached a definitive agreement to acquire data center operator Core Scientific in an all-stock transaction valued at approximately $9 billion. The deal marks one of the largest consolidations in the rapidly expanding AI infrastructure market, signaling a shift in how cloud providers secure the physical resources necessary to train and run next-generation AI models. By purchasing its primary infrastructure partner, CoreWeave is transitioning from a tenant to a landlord, fundamentally altering its operational economics.[1][2][3][4][5]

Under the terms of the agreement, which is expected to close in the fourth quarter, Core Scientific stockholders will receive 0.1235 newly issued shares of CoreWeave Class A common stock for each share they hold. This exchange ratio values Core Scientific at $20.40 per share—a 66 percent premium over its unaffected closing price in late June, prior to reports of advanced talks. Following the close, Core Scientific shareholders will own less than 10 percent of the combined entity.[1][2][3]

The primary driver behind the $9 billion price tag is not software or proprietary technology, but raw electrical power. The acquisition grants CoreWeave direct ownership of approximately 1.3 gigawatts of contracted gross power across Core Scientific’s national footprint of data centers, with an additional 1 gigawatt available for future expansion. To put that scale into perspective, a single gigawatt is roughly equivalent to the output of a standard nuclear reactor, capable of powering 750,000 homes.[1][5][6][7]

This massive energy footprint is essential because AI workloads—specifically the training of large language models and the subsequent "inference" process of generating responses—are exceptionally power-hungry. Unlike traditional web hosting, which relies on standard central processing units (CPUs), AI requires dense clusters of graphics processing units (GPUs). These specialized chips draw significantly more electricity and generate immense heat, requiring specialized cooling infrastructure that older data centers simply cannot support.[4][6]

Core Scientific’s journey to becoming a $9 billion AI infrastructure linchpin is a story of dramatic reinvention. The company originally built its massive facilities to mine Bitcoin, using energy-intensive computers to mint the cryptocurrency. When digital asset prices collapsed in 2022, Core Scientific was forced into bankruptcy. However, it emerged with a new strategy: repurposing its high-capacity power contracts and sprawling facilities to host AI workloads instead of crypto miners.[2][7]

The mechanics of this pivot are highly synergistic. Cryptocurrency mining facilities were intentionally built in regions with abundant, cheap electricity and were designed to handle extreme power densities. While the computing hardware itself is different, the underlying electrical substations, heavy-duty transformers, and industrial cooling systems required for crypto mining are exactly what AI cloud providers like CoreWeave desperately need.[4][7]

For CoreWeave, the financial logic of buying rather than renting is compelling. Prior to the acquisition, CoreWeave was Core Scientific’s largest tenant, having signed a series of 12-year contracts to host its GPU clusters. By acquiring the facilities outright, CoreWeave expects to eliminate more than $10 billion in cumulative future lease costs. Furthermore, the company projects $500 million in annual run-rate cost savings by the end of 2027 through streamlined operations and the removal of lease overhead.[1][2][3][5][6]

The deal represents a strategic shift toward "vertical integration" in the AI cloud sector. Historically, specialized cloud providers rented space from massive data center real estate investment trusts (REITs) and focused purely on managing the servers and software layer. By moving "down the stack" to own the physical buildings and power contracts, CoreWeave gains greater financing flexibility and absolute control over its expansion timelines, insulating itself from a tightening commercial real estate market.[1][4][8]

This aggressive expansion follows CoreWeave’s meteoric rise over the past three years. Backed by early investments from Nvidia and Microsoft, the company positioned itself as a nimble, GPU-centric alternative to hyperscalers like Amazon Web Services and Google Cloud. Following a highly successful initial public offering, CoreWeave has used its newly public stock as currency to accelerate its physical footprint, reporting staggering revenue growth of over 400 percent year-over-year in recent quarters.[1][3][5]

The broader data center market is currently experiencing unprecedented constraints. Vacancy rates in major North American markets have plunged to historic lows as AI adoption accelerates. Market research firm IDC projects that global spending on AI infrastructure will exceed $200 billion over the next five years, creating a severe bottleneck where the supply of power and physical space cannot keep pace with the demand for computing.[4][5]

Given this scarcity, the consolidation of critical infrastructure is beginning to attract regulatory attention. While CoreWeave is not as large as the traditional tech giants, its rapid acquisition of power-intensive data centers could draw scrutiny from antitrust enforcers at the Federal Trade Commission and the Department of Justice. Regulators are increasingly monitoring vertical mergers in the AI space to ensure they do not create insurmountable barriers to entry for smaller competitors seeking access to compute resources.[8]

An open question remains regarding the fate of Core Scientific’s legacy operations. While the company had pivoted heavily toward AI hosting, it still maintained a significant digital asset mining business. CoreWeave executives have indicated that they may "repurpose or divest" the remaining cryptocurrency mining operations over the medium term, fully dedicating the acquired gigawatts to enterprise AI workloads.[3][5]

For enterprise customers and AI developers, the acquisition promises greater stability. By owning its infrastructure, CoreWeave can theoretically deploy new GPU clusters faster and potentially offer more competitive pricing by cutting out the data center middleman. However, integrating two distinct corporate cultures—one rooted in agile cloud software and the other in industrial-scale crypto mining—presents significant execution risks.[4][6]

Ultimately, the $9 billion acquisition underscores a fundamental truth about the artificial intelligence revolution: it is bound by the laws of physics. No matter how advanced the algorithms become, they require vast amounts of electricity, specialized cooling, and physical real estate. By securing 1.3 gigawatts of power, CoreWeave is ensuring it has the foundational resources to power the next generation of AI development.[1][4]