Nvidia’s Huang Exit Scenario Points to a $16.8 Billion Terafab

By Rebecca Stone

About $1.9 trillion came out of Nvidia’s market value after its October 6 record, leaving the chipmaker worth about $3.9 trillion. At 7:10 a.m. ET on November 12, Nvidia shares traded at $159.98 in premarket trading, down 7.9% from Wednesday’s close and 33% from the peak.

By Thursday morning, Nvidia had announced no decision. The Wall Street Journal reported Wednesday evening that the board was preparing to oust Huang and could meet as soon as that night.

Huang had invited the scrutiny. He told employees days earlier that he had been unfaithful to his wife, after reports linked him to Alice Huang, a former executive at Singapore’s Megaspeed whom U.S. officials suspect of spying for China.

China calls that claim groundless, and Huang has not been accused of a crime. On Wednesday, The New York Times reported a second relationship, with Kimberly Powell, an Nvidia vice president who reports to him.

At 6:22 p.m. ET Wednesday, the market-commentary account Zero Hedge posted on X: “Elon please hire Jensen.” Two hours later, Elon Musk, described as the world’s richest man, replied: “That would be an honor.”

That was the exchange. No offer, role or commitment followed. It did, however, sharpen a conditional question: if Nvidia’s board removes Huang, what does a founder with both motive and means do next?

Where might Huang go next?

Huang co-founded Nvidia in 1993 and has run it ever since. Over those 33 years, Nvidia’s chips became the engine of the AI boom; after this week’s selloff, the company remained worth about $3.9 trillion.

The company’s name comes from invidia, the Latin word for envy. Co-founder Curtis Priem said the idea was to make competitors green with envy, and Nvidia’s green branding has carried the story since.

Huang holds about 811.5 million Nvidia shares directly and through trusts, according to his latest Form 4 filing. At the premarket price, those holdings were worth roughly $130 billion.

California, where Nvidia is based, generally does not enforce non-compete agreements, so Huang would face relatively few legal barriers in joining a rival effort. He has announced no plans. A Huang-Musk partnership would also test two chief executives accustomed to running companies their own way.

What is Terafab, and how large is it?

Musk announced Terafab on March 21 in Austin, Texas. Tesla and SpaceX own the chip manufacturing complex, which would make processors for Musk’s companies. Intel joined the project in April, and Musk said Terafab plans to use Intel’s 14A manufacturing process.

The stated long-term goal is one terawatt of AI computing hardware a year, a scale no chipmaker has attempted. It is a roadmap target, not demonstrated output or confirmed operating capacity.

The project has two parts. Musk put a smaller research fab near Tesla’s Gigafactory in Austin at about $3 billion; it would test ideas. Terafab’s permanent plant is planned for Grimes County, Texas.

County filings in May estimated $55 billion for the first phases and up to $119 billion for the full build. Those numbers came from filings rather than a disclosed funding commitment.

In August, Tesla and SpaceX said they would put about $16.8 billion into a first phase, with construction scheduled from December 2026 through 2028. The planned chips would serve Tesla’s cars and Optimus robots, along with hardware for SpaceX.

Why would Huang fit at Terafab?

Terafab is a manufacturing project. Nvidia designs chips and Taiwan Semiconductor Manufacturing Co. makes them; running a fab would put Huang on new ground.

The overlap goes beyond manufacturing. Self-driving cars and robots are the two industries Huang has praised most often as AI’s next wave, and Nvidia has spent years building chips and software for both.

Terafab would also give Huang a hand in shaping the semiconductor industry and a direct line into the autonomous-driving and robotics work he has championed.

For Musk, Terafab’s hardest problem is designing chips good enough to justify a fab of that size, and Huang’s record at Nvidia makes him a plausible technical leader.

Musk has said Terafab is needed because outside suppliers cannot make enough chips for his companies. Tesla and xAI also rank among Nvidia’s large hardware buyers, which sharpens the competitive risk for Nvidia.

What would this mean for shareholders?

For Nvidia holders, a founder leaving creates risk on its own. A founder leaving to help a customer build its own chips creates a larger one.

Dan Ives of Yorkville Ives said Nvidia would be worth 25% to 30% less without Huang. That is an analyst judgment, not a transaction value or company forecast.

For Tesla holders, with the stock at $388.70 before Thursday’s open, a Huang hire would be a vote of confidence in a project that still must prove it can be built on time and on budget.

Nothing about the scenario is settled. The board had not decided, Huang had said nothing about his plans, and Musk’s response was one sentence. Investors in either stock should know how much of their portfolio rides on the two men, including exposure inside index funds as well as direct holdings.

The next checkpoints are a Nvidia board decision, Terafab’s ability to deploy the disclosed $16.8 billion and start construction in December 2026, and whether the X exchange develops into a defined role. Until then, the project’s 1-terawatt annual goal is a roadmap, not output.