Nvidia invested $2 billion in CoreWeave through purchase of Class A common stock at $87.20 per share, nearly doubling its stake in the AI infrastructure provider as both companies deepen their strategic partnership around data center expansion. The transaction values CoreWeave shares below Friday's closing price of $92.98, with CoreWeave stock jumping 12 percent in Monday trading following the announcement.
The investment accelerates CoreWeave's ambitious goal to build more than 5 gigawatts of AI data center capacity by 2030, addressing critical infrastructure bottlenecks including land acquisition, power procurement, and physical facility construction. CoreWeave CEO Michael Intrator stated the deal enables the company to diversify its customer base while reducing dependency on any single client as it scales new data center capacity across global markets.
Nvidia's purchase adds approximately 23 million shares to its position according to Reuters calculations based on LSEG data, increasing its ownership from 6.3 percent to an estimated 13 percent and making Nvidia CoreWeave's second-largest shareholder behind Magnetar Capital. Before this transaction, Nvidia held approximately 24.3 million shares, positioning it as CoreWeave's third-largest shareholder.
The expanded collaboration extends beyond equity investment into product roadmap alignment, with CoreWeave gaining early access to deploy forthcoming Nvidia architectures including Rubin GPUs, Vera CPUs, and BlueField networking and storage systems. The companies will collaborate on testing and validating CoreWeave's AI-native software stack including its Mission Control resource-scheduling platform and SUNK infrastructure tools, aiming for improved interoperability with Nvidia's reference architectures for cloud and enterprise customers.
CoreWeave founded in 2017 as cryptocurrency mining company Atlantic Crypto but pivoted to cloud computing and AI infrastructure following the 2018 crypto crash. The company went public on Nasdaq in March 2025, debuting at $40 per share and quickly establishing itself as a specialized "neocloud" provider focused on GPU-intensive AI workloads. Major customers now include Microsoft as its largest client, alongside Meta, IBM, Cohere, OpenAI, and other AI model developers requiring massive compute capacity.
The investment follows CoreWeave's September disclosure of a $6.3 billion agreement with Nvidia to purchase unsold computing capacity through 2032, plus expanded partnerships including a $22.4 billion contract with OpenAI and $14.2 billion agreement with Meta for AI cloud infrastructure. These multi-billion dollar commitments provide CoreWeave with financial certainty to secure scarce resources where power, land, and grid access have become the primary bottlenecks constraining AI infrastructure growth rather than chip availability.
Nvidia's deepening financial ties with AI infrastructure providers have drawn scrutiny from investors concerned about circular financing dynamics, where the chip supplier simultaneously invests in companies that purchase its processors. A CoreWeave spokesperson clarified to Reuters that the new capital will not be used to purchase Nvidia processors, instead directing funds toward data center expansion, research and development, and workforce scaling.
The transaction reflects Nvidia's strategy of using capital and product roadmaps to shape where its next-generation platforms will be deployed, effectively building leverage across the AI infrastructure supply chain. Nvidia CEO Jensen Huang stated during a Monday interview that demand remains extraordinary as the industry enters early stages of AI infrastructure buildout, with power and facilities rather than chips becoming the binding constraint on growth.
CoreWeave's rapid ascent demonstrates how specialized cloud providers filling gaps left by traditional hyperscalers can capture enormous value. The company's "Nvidia-native" approach enables seamless integration that may be difficult for hyperscalers supporting diverse legacy hardware to match. However, the move creates challenging competitive dynamics for AMD and Intel attempting to penetrate high-end specialized cloud markets if Nvidia continues using its balance sheet to lock up promising infrastructure partners.
