Nvidia invests $2 billion in CoreWeave, expands partnership to scale AI data centers
Nvidia said it has invested $2 billion in CoreWeave at $87.20 per share and broadened its collaboration with the AI-focused cloud provider. The companies said the money and partnership are aimed at speeding CoreWeave’s buildout of large-scale AI data center capacity through 2030.

A major chipmaker backs an AI cloud specialist
Nvidia has invested $2 billion in CoreWeave and expanded the companies’ partnership as demand for AI computing infrastructure continues to rise. The investment was made at a purchase price of $87.20 per share, and CoreWeave stock jumped in premarket trading after the announcement.

CoreWeave has positioned itself as a specialized cloud provider focused on GPU-heavy workloads used to train and run AI systems. Nvidia, whose chips dominate the AI accelerator market, is increasingly tying capital and strategic partnerships to where its hardware platforms get deployed at scale.
Goal: faster land-and-power procurement for data centers
The companies said CoreWeave is targeting more than 5 gigawatts of AI data center capacity by 2030, and the fresh funding is intended to speed procurement of land and power needed to build facilities. Those two inputs—real estate and electricity—have become major bottlenecks as AI data centers expand.
Industry observers note that “neocloud” providers like CoreWeave have benefited from a surge in enterprise AI adoption, as companies seek quick access to accelerated compute without waiting for hyperscalers’ long build cycles.
What the expanded relationship includes
Alongside the investment, Nvidia and CoreWeave said they are deepening infrastructure and platform alignment to support the rollout of AI-focused facilities. The companies described the partnership as a way to meet rapidly growing compute demand while coordinating hardware, software, and deployment plans.
For customers, the stakes are straightforward: more capacity and smoother deployment can translate into faster model training timelines and more stable availability for AI inference at scale.