Overview
Nvidia has pledged a $2 billion equity investment in cloud‑compute specialist CoreWeave, cementing a partnership that spans hardware, software and financing.
Financial Commitment
The direct equity purchase links capital availability to hardware rollout schedules, ensuring that funding follows compute deployment milestones.
- $2 billion equity infusion
- Target: more than 5 GW of AI‑factory capacity by 2030
- Joint focus on accelerating AI‑factory construction
Multi‑Generation Hardware Strategy
CoreWeave will deploy several generations of Nvidia platforms, including early adoption of the Vera Rubin system, Vera CPUs and BlueField storage solutions.
- Vera Rubin GPUs delivering up to 28.8 exaflops with 576 GPUs
- Custom Arm‑based Vera CPUs with high core counts and large coherent memory
- BlueField smart NICs for high‑bandwidth interconnects
AI Factories & Power Scaling
The collaboration aims to build “AI factories” that combine massive compute with reliable power, leveraging Nvidia’s AI‑optimized hardware and CoreWeave’s cloud expertise.
- Projected >5 GW of dedicated AI compute power by 2030
- Integration of HBM4 memory modules from Samsung for next‑gen performance
- Potential inclusion of AWS in Nvidia’s AI‑factory ecosystem
Industry Impact
This partnership illustrates two key market narratives: the growing importance of server‑grade CPUs in AI workloads and the shift toward modular, high‑end CPUs as cost‑effective alternatives to full rack‑scale systems.
- Supply‑chain pressure on high‑performance CPUs
- Lower entry costs for specialized AI deployments
- Nvidia uses CoreWeave as a proving ground for full‑stack AI solutions