Nvidia has crossed $40 billion in equity commitments this year, transforming from chip supplier into the primary financier of the AI supply chain. The strategy is straightforward: fund the companies that build, power, and deploy AI infrastructure, ensuring it all runs on Nvidia hardware.
Last week alone, Nvidia secured investment rights worth $2.1 billion in data center operator IREN and $3.2 billion in glass manufacturer Corning. IREN will deploy up to 5 gigawatts of Nvidia's DSX infrastructure designs globally. Corning is building three U.S. facilities dedicated to optical technologies for Nvidia's rack-scale systems, signaling a shift from copper to fiber interconnects.
These moves follow a pattern. Nvidia invested $2 billion each in Marvell Technology, Lumentum, and Coherent for silicon photonics development. It poured another $2 billion into CoreWeave and Nebius Group, the so-called neocloud operators building data centers around Nvidia's architecture. The single largest bet: $30 billion into OpenAI.
From a systems architecture perspective, Nvidia is vertically integrating the entire AI stack through financial instruments rather than acquisitions. The company's balance sheet now carries $22.25 billion in non-marketable equity securities, up from $3.39 billion a year ago. Investment gains hit $8.92 billion, largely from its Intel stake, which quintupled in months.
The engineering trade-off is becoming visible. Nvidia generated $97 billion in free cash flow last fiscal year. Deploying that capital into partners creates a feedback loop: companies get funding to build capacity, which requires Nvidia GPUs, which generates more revenue. Critics call it vendor financing reminiscent of the dot-com era. Wedbush analyst Matthew Bryson describes it as a "circular investment theme" that builds a competitive moat if executed correctly.
Mizuho's Jordan Klein sees the component investments as smart capital allocation, accelerating development of scarce technology. He's less convinced about neocloud deals, which "smell like pre-funding the purchase of your own GPUs." The risk, as Creative Strategies' Ben Bajarin notes, is that when the cycle turns, it becomes unclear how much demand was organic versus balance-sheet-supported.
Nvidia's own SEC filing states these investments target early-stage startups and AI model companies that purchase its products directly or through cloud providers. CEO Jensen Huang, on an April podcast, said the company tries to invest in all major foundation model companies rather than picking winners.
With Q1 earnings approaching, shareholders will see how this portfolio affects financials. The $30 billion OpenAI deal may be the last pre-IPO check Nvidia writes. For now, the company is building an infrastructure ecosystem where every layer—glass, photonics, data centers, models—depends on Nvidia silicon.
Source: CNBC