Nvidia is drawing market attention as it aggressively expands equity investments across the global artificial intelligence (AI) ecosystem, backed by its massive cash-generating capacity.
Nvidia has made equity investments of about $40 billion in AI infrastructure this year, U.S. business network CNBC reported Tuesday. The investments span the AI value chain, from model developers such as OpenAI to optical fiber manufacturers and emerging cloud companies.
Related valuation gains are also surging along with the expanded investments. The value of Nvidia's private equity investments stood at $22.25 billion as of the end of January, more than six times the $3.39 billion recorded a year earlier, according to the company's financial statements. Valuation gains on listed equities reached $8.92 billion, driven by returns on its Intel stake, among others.
The leading names in AI models are all on Nvidia's investment list. In addition to a $30 billion investment in OpenAI, Nvidia has backed Anthropic and xAI. The strategy is interpreted as an effort to broaden the market by investing across major AI model developers rather than betting on a single company, ultimately expanding demand for its own hardware.
Nvidia Chief Executive Officer Jensen Huang has long described the AI industry as a "five-layer cake" structure running through energy, chips, infrastructure, models and applications, stressing the importance of the entire ecosystem.
Nvidia is strengthening the infrastructure layer through its investment in data center operator Iren, and securing the model layer through investments in foundation model companies such as OpenAI. At the same time, it is extending its influence into networking technology with stakes in glass and optical fiber maker Corning and optical technology firms Marvell and Lumentum.
The market is paying particular attention to Nvidia's aggressive investments in "neo-cloud" companies such as CoreWeave and Nebius. These are AI-focused cloud companies that deploy Nvidia graphics processing units (GPUs) at scale. For Nvidia, the approach reduces reliance on major Big Tech clients while building an AI infrastructure network centered on its own chips.
Some in the investment community, however, point out that the structure effectively takes the form of "circular transactions." Companies that receive investment from Nvidia use the funds to purchase Nvidia GPUs, creating a self-reinforcing flow.
"There are aspects that match the circular transaction structures of the dot-com bubble era," Matthew Bryson, an analyst at Wedbush Securities, said. Jordan Klein, semiconductor analyst at Mizuho, also raised questions about the concentrated investments in neo-cloud companies.
Analysts suggest that the $97 billion in free cash flow Nvidia generated in its most recent fiscal year is flowing back into the market and returning as orders for Nvidia products, leaving room for the pattern to be interpreted as a bubble.
Nvidia counters that the investments are aimed at nurturing the AI ecosystem rather than being simple financial transactions. "We should not pick winners; we should support everyone," Huang said in an interview last month. At the company's February earnings call, he also stressed that "our investments are focused on expanding and deepening the reach of the ecosystem."
