Nvidia Backs AI Labs Driving a Quarter of Its Upcoming Revenue

Nvidia expects that artificial intelligence labs supported directly or indirectly through its own balance sheet will drive approximately 25% of its revenue in the coming year. The hardware giant has invested nearly $50 billion into AI developers and helped organize frameworks for more than $500 billion in external capital commitments.

The Capital and Compute Feedback Loop

During a recent earnings call, Nvidia Chief Financial Officer Colette Kress addressed the financial mechanics linking Nvidia’s investments to hardware sales. Under this structure, Nvidia invests in AI startups or secures credit arrangements, which the labs subsequently utilize to build computing infrastructure filled with Nvidia GPUs. Revenue from those chip sales flows back to Nvidia, bolstering its cash reserves and enabling further investments.

To scale this ecosystem, Nvidia has formed preliminary alliances with major financial institutions—including Apollo Global Management, BlackRock, Blackstone, Brookfield Asset Management, Goldman Sachs, and KKR—aiming to pool more than $500 billion in outside funding. Binding contracts for these arrangements remain subject to definitive agreements.

Major Infrastructure Commitments

Beyond venture investments, Nvidia is securing essential data center assets directly:

• Power and Facility Access: Through a partnership with SB Energy, Nvidia secured infrastructure designed exclusively for its hardware. The initial 4.25-gigawatt phase is dedicated to OpenAI, which has projected compute needs reaching roughly 12 gigawatts through 2030.
• Credit Support: Nvidia is backing nearly two gigawatts of infrastructure capacity for an undisclosed AI research lab.
• Cloud Capacity Backstops: Nvidia assists smaller cloud operators by guaranteeing to rent unused capacity, providing lenders with predictable revenue while taking a share of upside earnings.

Pushback Against ‘Circular Financing’ Concerns

Addressing market scrutiny over potential circular financing risks, Kress argued that third-party financiers independently assess the risk profile of each deal. She emphasized that Nvidia is not directly issuing loans and that underlying hardware can readily be reallocated to other buyers if any single client defaults.

"Young AI labs face unprecedented computational demand that surpasses their existing balance sheets, making access to compute the primary bottleneck to growth rather than market demand," Kress explained.

Surge in Agentic AI Demand

Driving this massive buildout is the rapid transition toward autonomous AI agents. According to Nvidia leadership, agentic workflows require between 15 and 100 times more processing power than human-prompted queries.

Backed by this sustained demand, Nvidia issued quarterly revenue guidance of $108 billion and projected approximately 70% growth heading into 2028. However, rising memory component costs are expected to slightly compress gross margins from 74% toward the 71%–72% range in subsequent quarters.