Nvidia names interim CEO amid Taiwan packaging expansion

Illustration for: Nvidia names interim CEO amid Taiwan packaging expansion
AI-generated illustration. Visual interpretation does not represent real individuals or scenes.

THE BARE STORY

Nvidia has named Chief Financial Officer Colette Kress as interim chief executive after co-founder Jensen Huang resigned as president, CEO and board member, the company announced Thursday. The change comes amid developments in Taiwan’s semiconductor sector.

Siliconware Precision Industries said it had reached a multibillion-dollar agreement with Nvidia to expand advanced chip-packaging capacity in Taiwan. The company plans to triple that capacity by the middle of 2029.

Industry executives and investment advisers said Huang’s departure was unlikely to significantly disrupt Taiwan’s semiconductor supply chain, citing Nvidia’s CUDA software platform and lengthy supplier approval processes. They also noted concerns about U.S. export rules for China.

Same Facts. Different Perspectives.

Two AI models. Two viewpoints. One factual foundation.

The headline is a personnel announcement. The decision is a map.

Nvidia can name Colette Kress, its longtime chief financial officer, interim chief executive, and the machines will keep moving. That is what the comforting chorus means when it says Jensen Huang’s resignation will not disrupt Taiwan’s semiconductor chain — CUDA lock-in, years-long supplier qualifications, a company bigger than one man. They are not wrong about succession. They are wrong about what succession is being asked to hide.

The same moment the board changed the name on the door, Siliconware Precision Industries said it had a multibillion-dollar agreement to expand advanced packaging for Nvidia in Taiwan, with that capacity set to triple by the middle of 2029. Packaging is not a footnote under the fab. For the accelerators that train and run the systems Washington now treats as strategic, advanced packaging has been the choke point — the step that turns a leading-edge die into something that actually computes. Nvidia is not hedging that choke point. It is enlarging it, on the island Beijing refuses to renounce seizing and has repeatedly rehearsed isolating.

There is a serious version of the silicon-shield argument: a democratic ally the world cannot afford to lose is harder to attack. There is also the lazy version, and that is the one corporate planning and political habit have settled into. Put more of the irreplaceable work in the most exposed place, then call the exposure deterrence. Deterrence is a military and political fact. It is not a logistics plan. Software does not cross a strait under blockade. A qualification cycle measured in years is a reason to move capacity now, not a press release about how safely nothing can change.

The other constraint in this story, mentioned almost as an aside, is U.S. export rules for China. Those rules exist because advanced chips are not consumer gadgets. Feeding the frontier to a rival’s military-civil industrial base is not free trade. It is subsidizing the threat you then ask sailors and airmen to contain. Here the Right has to be honest about the incentive it creates. Make China a worse customer and Taiwan looks even better to a firm optimizing for the next several quarters. Controls without a hard push to put advanced packaging, and the process technology around it, in the United States and in allies that are not the central object of Chinese war planning do not produce resilience. They produce a richer bullseye.

That is where a conservative argument has to be willing to irritate its own donors. Too many on the Right have treated Nvidia’s valuation as proof that America already won the future, and treated any limit on China sales as an assault on shareholders. That is not nationalism. It is a stock tip with a flag on it. Markets allocate capital brilliantly inside a stable order. They are not a China policy. If the interim CEO’s job is to protect margins by keeping the 2029 plan parked in Taiwan and lobbying to sand the edges off the export rules, the board will have served its owners and left the country holding a risk no earnings call can hedge.

None of this requires pretending Washington is good at building factories. The CHIPS era already showed how slowly money becomes clean rooms, and how eagerly industrial policy picks up unrelated political freight. Clumsy government is not an argument for strategic surrender. Judge the project by the only number that matters: how much of the packaging Nvidia cannot live without sits outside a Chinese missile envelope, on a clock shorter than Beijing’s. Tripling Taiwan capacity by 2029 is an industry fact. It is also an indictment of how little alternative the United States has forced into existence.

Huang built an extraordinary company. If the technicians are right, his exit will not stop the wafers. That is exactly why the soothing story fails. The danger was never that Nvidia would forget how to design a chip. The danger is that the physical power to build the AI age stays concentrated on an island China has already decided may not remain separate — and that American policy keeps mistaking that concentration for an alliance strategy.

Kress can run the numbers. Washington has to decide whether 2029 is a milestone or a deadline.

How it may affect me

If your retirement account, employer, or AI plans run through Nvidia, the risk through 2029 is not the interim CEO. It is Taiwan packaging concentration plus U.S. export rules on China — either can reprice the stock, delay hardware, and raise computing costs faster than a leadership handoff. Treat a heavy bet on that supply chain as a geopolitical position, not a neutral holding.

Read the story at

Note: All TheBareNews content is AI-generated. For additional context, reporting, and updates, you are invited to explore the news outlets linked above.

Stories You May Have Missed