Nvidia has crossed a striking milestone in its evolution from chip vendor to AI financier: more than $40 billion committed to equity investments across the AI stack so far in 2026, according to reports published by CNBC and TechCrunch on May 9. The figure recasts the world's most valuable chipmaker as one of the year's most active strategic investors, and it has reignited a debate about whether Nvidia is building a durable ecosystem or quietly bankrolling its own demand.
A Vendor Turned Investor
The headline number is anchored by a $30 billion commitment to OpenAI, the single largest equity check Nvidia has ever written. That stake sits inside a broader strategic partnership in which Nvidia has signaled it could invest up to $100 billion progressively as each gigawatt of OpenAI compute is deployed.
Beyond OpenAI, Nvidia's deal cadence has accelerated this spring. In the past week alone the company struck an agreement giving it the right to invest up to $2.1 billion in data-center operator IREN, a day after a pact allowing up to $3.2 billion into 175-year-old glassmaker Corning to expand US optical-fiber capacity for AI infrastructure. Reports also point to roughly two dozen private startup rounds this year, on top of 67 venture deals tallied in 2025.
The scale is visible on the balance sheet. Nvidia's non-marketable equity securities swelled to $22.25 billion at the end of January from $3.39 billion a year earlier, reflecting how aggressively the company has been buying into the supply chain it powers.
Ecosystem Strategy or Vendor Financing?
CEO Jensen Huang has framed the strategy in ecosystem terms. "Our investments are focused very squarely, strategically on expanding and deepening our ecosystem reach," Huang told investors on Nvidia's last earnings call in February. He has separately argued that Nvidia tries to back "all" of the leading foundation-model labs rather than picking winners.
Not everyone reads it that way. Wedbush Securities analyst Matthew Bryson said this week's deals fall "squarely into the circular investment theme" that has been driving fears about the durability of the AI build-out. The worry is that Nvidia equity is funding companies that then route those dollars back through Nvidia hardware orders — a pattern critics liken to the vendor financing that helped inflate the late-1990s telecom bubble.
Implications
For the AI industry, the immediate effect is more capacity coming online faster: data centers, optical interconnects, and frontier-lab compute commitments that would be harder to finance through conventional channels alone. For investors, the question is harder. If demand for AI inference and training keeps compounding, Nvidia's equity stakes could prove to be the most lucrative side bet in modern corporate history. If growth slows, the same web of cross-holdings could amplify a downturn the way it currently amplifies the upswing. Either way, May 9's $40 billion tally makes clear that Nvidia's role in the AI economy is no longer limited to selling silicon.
