Nvidia's US$3 billion investment in Meta's share sale, announced on October 27, is the latest addition to a portfolio of stakes, loans and guarantees that has made the chip giant what some analysts describe as the central bank of the artificial intelligence boom.
The deals have kept demand high for Nvidia's processors, which power the data centres at the heart of the AI revolution, making Nvidia the world's most valuable company and chief executive Jensen Huang AI's kingmaker.
But the investments have raised a nagging question among some doubters on Wall Street: is Nvidia helping to inflate an AI bubble?
Stakes across the whole AI stack
Nvidia has taken stakes in leading AI developers including OpenAI, Anthropic and Elon Musk's xAI, now part of SpaceX. It has also invested in "neocloud" companies such as CoreWeave and Nebius, which rent out computing power built on Nvidia chips.
Its bets extend well beyond the United States. In Europe, Nvidia has backed French AI developer Mistral and British self-driving startup Wayve. In Asia, it has teamed up with South Korea's SK Group to build data centres.
Beyond equity, the company has offered financial backstops, acting as a customer of last resort. In CoreWeave's case, Nvidia committed to buy up to US$6.3 billion of unused computing capacity if no other customer takes it. Nvidia also has around US$36 billion in cloud backstops and a guarantee of up to US$105 billion linked to an OpenAI data centre, according to US media reports.
Why Nvidia is playing banker
Experts say Nvidia's enormous profits give it few better options. "They're essentially becoming a bank," said Michael Cusumano, a professor at the MIT Sloan School of Management. "Channeling it back to infrastructure investment tied to purchases of Nvidia equipment is a smart thing."
Running its own data centres would put Nvidia in competition with its biggest customers "and that might also trigger a whole bunch of antitrust complaints," Cusumano added.
Gil Luria, head of technology research at D.A. Davidson, said Nvidia spots parts of the AI ecosystem that are underinvesting and funds them to fill the gaps. "It's unprecedented, but it appears to be working," he said.
Huang has defended the approach, saying frontier AI labs need tens of billions to get started, not millions.
The circular-financing worry
Critics have compared the deals to the dot-com era, when firms inflated growth by investing in startups that bought their products. "There is nothing intrinsically wrong with Nvidia helping its clients," said Paulo Carvão, a senior fellow at the Harvard Kennedy School. "But the closer these relationships are, the harder it is for one to assess true sustainable demand."
Luria said not all demand was equal. "If Nvidia gives CoreWeave money to turn around and buy Nvidia chips, that's not as good as Microsoft buying Nvidia chips because they have a customer in OpenAI that has a customer in an enterprise technology that's willing to pay for it," he said.
The main concern is concentration: two direct customers made up 22 percent and 14 percent of Nvidia's fiscal 2026 revenue, per its annual report. Carvão warned the overlapping investments raise contagion risk if a weaker firm such as CoreWeave stumbles. "If AI demand disappoints, spending cuts could hit chipmakers, data-centre builders and their financiers at once," he said.
Shares of Nvidia-linked neocloud companies including CoreWeave, Nebius and IREN fell at least 30 percent in October amid worries about tighter credit for AI projects. Carvão put continued demand growth at 12 to 18 months; Cusumano said three to five years.
