JupiteX Get the app
Economy & Business18 Nov 2026 · about 7 min

How Jensen Huang’s exit from Nvidia impacts China’s AI supply chain

The brief

Nvidia’s board has appointed chief financial officer Colette Kress as interim chief executive after Jensen Huang resigned. The temporary appointment gives the company a leader while it searches for a permanent successor. Huang’s departure matters because Nvidia is the world’s top chipmaker and plays several roles in the AI industry: hardware provider, investor, and community leader. The board’s stated plan is straightforward: Kress will serve as chief executive on an interim basis, and the company will later name a permanent successor. The article does not give a timetable for that decision or describe changes to Kress’s responsibilities as chief financial officer. For now, Kress is leading Nvidia during a period of uncertainty. Huang’s exit has raised questions about Nvidia’s future and its relationships across the AI industry. Executives at Taiwanese suppliers nevertheless told Taiwan’s Central News Agency they expected Nvidia orders to remain steady in coming years, citing sustained AI demand.

01

Who is leading Nvidia while it searches for a permanent successor to Jensen Huang?

Nvidia’s board has appointed chief financial officer Colette Kress as interim chief executive after Jensen Huang resigned. The temporary appointment gives the company a leader while it searches for a permanent successor. Huang’s departure matters because Nvidia is the world’s top chipmaker and plays several roles in the AI industry: hardware provider, investor, and community leader.

The board’s stated plan is straightforward: Kress will serve as chief executive on an interim basis, and the company will later name a permanent successor. The article does not give a timetable for that decision or describe changes to Kress’s responsibilities as chief financial officer.

For now, Kress is leading Nvidia during a period of uncertainty. Huang’s exit has raised questions about Nvidia’s future and its relationships across the AI industry. Executives at Taiwanese suppliers nevertheless told Taiwan’s Central News Agency they expected Nvidia orders to remain steady in coming years, citing sustained AI demand.

02

What does the article say could happen to Nvidia chip access in China after the scandal?

The article says the scandal could lead the U.S. government to tighten controls on chip access. That matters because Nvidia chips remain important to China’s AI industry, even as Chinese chipmakers gain ground. The United States already bans exports of advanced Nvidia chips to China, and the article describes efforts to obtain them through other routes.

One example involves Alice Huang, an executive at Singapore-based data center company Megaspeed and one of Nvidia’s biggest customers in Southeast Asia. The Information reported that she helped move about 240,000 H100, H200, and some A100 chips into China. Chinese companies have also tried to access chips remotely through data centers in Southeast Asia.

The article reports that Chinese AI labs are trying to secure overseas supplies in anticipation of tighter controls, while traders in China are stockpiling restricted chips. It does not say exactly what new rules may be introduced. The possibility of a crackdown could make access more uncertain, while domestic alternatives still have performance and production limits.

03

In the first half of the year, what share of China’s AI accelerator market did Nvidia and Huawei each have?

Nvidia and Huawei together accounted for most of the China AI accelerator market in the first half of the year, according to figures cited in the article. Nvidia had a 34% share, compared with 28% for Huawei. These shares show both Nvidia’s continued presence and the growth of a major domestic competitor.

Nvidia shipped 678,000 AI accelerators in China during that period, down 44% from a year earlier. Huawei shipped 572,000, up from more than 300,000 a year earlier. Alibaba’s T-Head held 7% of the market, Cambricon 5%, Baidu subsidiary Kunlunxin 4%, and Hygon 3%.

Chinese chipmakers are gaining share, but the article says Nvidia chips remain essential to the AI industry. Most AI developers still train new models on Nvidia chips, and Huawei lags in both performance and production volume. The market shares describe shipments and sales in that first half, not a guarantee of future market positions.

04

What are export controls, and how do they restrict the sale of advanced chips to China?

In general, export controls are government rules that restrict the sale or transfer of specified goods to particular destinations or buyers. They matter here because advanced chips can be subject to restrictions, limiting which companies may sell them to China. The article says the United States has banned exports of advanced Nvidia chips to China.

A control can work through a ban or through a requirement to obtain government permission before exporting. In practice, a ban blocks covered sales, while a permission requirement allows authorities to review or deny a proposed shipment. The article does not describe the detailed rules or licensing process for Nvidia products.

The controls have not stopped every attempt to access chips: the article reports that Chinese companies have sought remote access through Southeast Asian data centers. It also says the U.S. government may tighten controls after the scandal. Chinese AI labs are trying to lock in overseas supplies, and traders are stockpiling restricted chips, according to the reports cited.

05

What Chinese-made chips could AI companies use instead of Nvidia chips, and what limits do those alternatives still face?

Chinese AI companies have domestic alternatives to Nvidia accelerators, including chips from Huawei, Alibaba’s T-Head, Cambricon, Baidu subsidiary Kunlunxin, and Hygon. These options matter because U.S. restrictions and possible tighter controls could make Nvidia chips harder to access. Chinese AI labs are also working with chipmakers to ensure their models can run on domestic infrastructure.

The market figures show these suppliers are already serving customers: Huawei held 28% of China’s AI accelerator market in the first half of the year, while T-Head had 7%, Cambricon 5%, Kunlunxin 4%, and Hygon 3%. But a growing market share does not mean the chips match Nvidia in every respect.

The article says most AI developers still train new models on Nvidia chips. Research firm Epoch AI estimates Nvidia’s B300 has 6.75 times the arithmetic performance of Huawei’s best Ascend 950 chip. Huawei also lags Nvidia in production volume. Domestic chips are catching up, but the article describes limits that could make replacing Nvidia difficult.

06

Why are companies in Taiwan and South Korea important to Nvidia’s supply chain?

Taiwanese and South Korean companies matter because the article says both countries host a vast network of Nvidia suppliers. Their role connects Nvidia to a wider regional supply chain, so questions about Nvidia’s future can affect companies beyond the chipmaker itself. The article does not identify specific suppliers or describe what each one provides.

Taiwan has a particularly visible relationship with Nvidia. Huang, who was born in Taiwan, visited its night markets, shared snacks with fans, and hosted banquets with executives of Taiwanese suppliers. He also said Nvidia planned to invest $150 billion a year in Taiwan. These details show how closely Nvidia’s public profile and business relationships were linked to the island.

Huang’s resignation has created uncertainty about Nvidia’s supply-chain relationships with Asian companies. Still, executives at Taiwanese suppliers told Taiwan’s Central News Agency that they expected steady Nvidia orders in the coming years, pointing to sustained demand from the AI industry. The article gives no comparable forecast from South Korean suppliers.

07

What is an AI accelerator, and why does training an AI model require so much computing power?

In general, an AI accelerator is a chip built to handle computing tasks used by artificial intelligence systems. The article discusses accelerators as chips supplied by Nvidia and Chinese companies, but it does not define the term. These chips matter because access to them affects how AI developers build and train models.

Training an AI model involves repeatedly processing data and adjusting the model through calculations. Those calculations can be demanding, which is why developers use specialized chips and often many of them. The article does not explain the technical process of training, but it notes that most AI developers still train new models on Nvidia chips.

The article compares Nvidia and Chinese-made accelerators by market share, shipments, performance, and production volume. Nvidia’s B300 delivers 6.75 times the arithmetic performance of Huawei’s best Ascend 950 chip, according to Epoch AI. Chinese chips are catching up, but the article says Nvidia remains essential and Huawei also lags in production volume.

This brief was written by AI from the original reporting and checked by other models. Names, figures and quotes come from the source; read it for full context.

Read more in the JupiteX app

Pulse is free. New stories every 4 hours, each one broken into the questions that explain it.

Or read more news on the web