Google Advertisement
Virginia, Oktober 21, 2026 – Nvidia backed data center developer Nscale is negotiating a new US$45 billion cloud computing agreement with AI lab Anthropic, according to sources familiar with the matter.
The deal, if finalized, would mark Anthropic’s second mega contract with Nscale in just two months, underscoring the extraordinary demand for advanced AI infrastructure even as investor concerns mount over debt fueled expansion.
The proposed agreement centers on Nscale’s Monarch campus in West Virginia, where a second facility is planned to deliver 460 megawatts of computing capacity powered by Nvidia’s latest Vera Rubin AI chips.
Anthropic, the developer of the Claude chatbot, is also negotiating a right of first refusal for a third Monarch building, signaling its long term reliance on Nscale’s infrastructure.
Google Advertisement
While discussions are ongoing and terms remain fluid, the scale of the talks highlights the intensity of competition for AI cloud resources.
Nscale’s meteoric rise has been fueled by Nvidia’s backing and massive client commitments.
In just two and a half years, its contracted value has surged from US$100 million to US$103 billion.
Anthropic’s initial US$45 billion rental deal in August was already one of the largest AI cloud agreements ever signed.
A second deal of equal size would cement Nscale’s position as a dominant player in the sector.
Yet the expansion comes amid growing unease in financial markets. Rival CoreWeave recently saw its shares tumble on concerns over a US$50 billion debt load, sparking a broader selloff in A Linked stocks.
Nscale itself faced setbacks when Microsoft withdrew from a flagship data center project due to power connection delays, pushing supply timelines into the 2030s.
Meanwhile, Bloomberg reported that investors have urged Anthropic to lower its IPO valuation target, reflecting skepticism about the sustainability of such aggressive growth.
For Anthropic, securing vast computing power is critical to scaling Claude and competing with rivals like OpenAI and Google DeepMind.
For Nscale, another mega-deal would reinforce its role as a leading AI infrastructure provider, though questions remain about the durability of its debt heavy model.
For Nvidia, the deployment of Vera Rubin chips further entrenches its dominance in AI hardware, even as hyperscalers develop in-house alternatives.
If completed, the deal would represent one of the largest commitments in the history of cloud computing.
But with debt risks, infrastructure delays, and volatile investor sentiment, the AI sector may be entering a more precarious phase nviwhere execution challenges loom as large as technological ambition.
