Nvidia isn’t just selling the engines; it’s funding the racetracks.
In a massive strategic deepening, Nvidia (NVDA) has invested another $2 billion into “neocloud” provider CoreWeave, becoming its second-largest shareholder. The move is designed to accelerate the physical build-out of data centers required to sustain the AI boom.
💰 THE DEAL METRICS:
- Investment: $2 billion via share purchase at $87.20/share.
- The Stake: Nvidia adds ~23 million shares, nearly doubling its previous position (6.3%) to cement its status as a top backer.
- The Target: CoreWeave is aiming to build over 5 Gigawatts of AI data center capacity by 2030.
🏗️ LAND, POWER, & SPEED: Crucially, this capital is allocated for physical constraints, not silicon.
- The Bottle Neck: The industry is no longer just chip-constrained; it is power-constrained.
- The Use of Funds: CoreWeave will use the cash to speed up the procurement of land and power access, as well as R&D and workforce scaling.
🔄 ADDRESSING THE “CIRCULAR” RISK: With regulators and investors scrutinizing “round-tripping” (where AI startups use investment cash to buy Nvidia chips, boosting Nvidia’s revenue), CoreWeave explicitly stated:
- The Defense: A spokesperson confirmed the new funds will not be used to purchase Nvidia processors, but are strictly for operational and physical infrastructure expansion.
💡 ANALYST TAKEAWAY: This is vertical integration by proxy. CoreWeave (a former crypto miner turned AI specialist) offers a nimbler, GPU-native alternative to the traditional Hyperscalers (AWS/Azure/Google). By supercharging CoreWeave’s balance sheet, Nvidia ensures there is a “friendly” massive-scale buyer in the market who prioritizes their architecture, keeping pressure on the legacy cloud giants to keep spending.
👇 Cloud Architects: Will “Neoclouds” like CoreWeave permanently capture the AI training market, or will the Hyperscalers eventually catch up on GPU density?