Nvidia Shares Fall 13 Percent as Jensen Huang Exit Hits AI's Central Bank

Nvidia shares fell 13 percent in early trading after Jensen Huang's departure, a plunge that measures exactly how much the world's second-most valuable company depended on one man. The drop raises a harder question: whether the AI economy can function without what industry observers call its "central bank."

Nvidia's valuation rests on a single fact. The company designs and sells the advanced AI chips — graphics processing units, or GPUs — that underpin the AI economy. Without them, models such as ChatGPT and Claude could not be trained and would not run. That position made Nvidia arguably the most important player in AI and, at times, the most valuable company in the world.

What does Huang's exit actually change?

The design of AI chips is not Nvidia's monopoly. Most leading technology companies, including Microsoft, Meta and Google, are building their own lines of AI processors. Some of these, notably Google's tensor processing units (TPUs), appear at least as good as Nvidia's chips and better suited to certain workloads.

What made Nvidia irreplaceable — and far more integral to the AI economy — was its web of relationships, and Huang personally led most of them.

The first strand is financial. Nvidia sits inside a complex mesh of ties with the companies building AI infrastructure, a circular-financing arrangement that implicates US technology giants and many of the world's governments.

The second strand is diplomatic. In 2025, Huang met numerous world leaders — including Mohammed bin Salman, Keir Starmer and Emmanuel Macron — and persuaded them of the urgency of developing "sovereign AI" built on his product. His presence served as a guarantee of those nations' AI plans. The meetings were followed by deals in which Nvidia announced arrangements to supply those governments with scarce, hard-to-obtain chips.

None of these agreements was particularly transparent. Nvidia appears to have promised 260,000 advanced chips to South Korea, for example, but it was never immediately apparent when the chips would be delivered, who would pay for them, or on what terms.

How deep does Nvidia's financing of its own customers run?

Nvidia financed a web of companies known as "neoclouds," which bought its chips, installed them in datacentres and rented capacity to the world's biggest AI companies. The arrangement served Nvidia well: the neoclouds propped up demand while keeping the risk of building datacentres and finding customers for them off Nvidia's balance sheets.

The neoclouds were not the whole of it. Nvidia earned its "central bank of AI" label because it became enmeshed in almost every layer of the AI economy. It invested in a spate of AI startups that could become future customers. It offered massive cash sums to companies such as OpenAI, to be used for purchasing its chips. The total investment runs to hundreds of billions of dollars — Nvidia's money, circling again and again through a limited set of accounts.

Why does TSMC matter more than any chip designer?

Nvidia's position also rests on its relationship with TSMC, the Taiwan Semiconductor Manufacturing Company, which manufactures the world's AI chips across several fabrication plants on the north-west side of Taiwan.

In geopolitical terms, TSMC arguably matters more than Nvidia or any other chip designer on earth. Companies other than Nvidia can design AI chips. TSMC manufactures the overwhelming majority of them and has no real competitor, because manufacturing AI chips is one of the most expensive and difficult industrial processes humanity has created.

Each chip is etched with nanometre-scale lines by lasers working in the extreme ultraviolet spectrum — machines that cost billions of dollars and took two decades of research to develop. TSMC has refined its own process for more than two decades, yet it still records only a 60 percent yield rate on its most advanced chips. Forty percent of the chips coming off its production line have defects. That is a staggeringly high rejection rate for one of the world's most expensive technologies.

Nvidia has worked extremely closely with TSMC and helped the manufacturer develop its processes and products. Company lore holds that Huang wrote TSMC's chief executive a personal letter of introduction in 1997, nearly 30 years ago. Other companies cannot match that. These are high-trust relationships built over decades.

What remains hidden in Nvidia's supply chain?

Whether any of this survives Huang intact is nearly impossible to answer, because Nvidia's relationships are so opaque. Investigators tracing the supply chain of its chips run into secretive corporate contracts and tangled export records. Customers are known to wait years in a queue for Nvidia's chips — but who is in that queue, in what order, and how many chips have actually shipped remain open questions.

The Megaspeed affair brought this murkiness into sharp focus. Megaspeed, a small Singaporean company, obtained access to large quantities of cutting-edge GPUs even as other companies struggled. One of its executives, Alice Huang, was later revealed to have been having an affair with Nvidia's chief executive and to be an alleged Chinese spy. The relationship appears to have allowed Megaspeed to secure roughly $7.5bn (£5.67bn) worth of Nvidia chips, which were allegedly smuggled into China.

Could a new chief executive bring transparency?

Perhaps Huang and his glad-handing were themselves a source of the opacity. A new chief executive might bring some transparency and predictability to chip contracts and supply chains.

But transparency cuts both ways. If the "central bank of AI" opened itself to external scrutiny, elaborate debt structures and handshake agreements buried in its contracts would have to be uncovered.

Huang dazzled the world in his leather jacket and convinced heads of government that his technology held the keys to transformation and to power. The 13 percent share plunge shows the market's first verdict on life without him; the audits, contracts and supply chains he leaves behind will determine the rest.