Nvidia's board announced on November 12 that Jensen Huang has resigned as president, chief executive officer and a member of the board, effective immediately. He had run the company for 33 years, taking it from a graphics card vendor to the supplier at the center of the AI buildout, and at one point a market value approaching $5 trillion.

The handover

Colette Kress, executive vice president and chief financial officer, has agreed to serve as interim CEO while keeping her CFO duties. She joined Nvidia in September 2013 after nearly 25 years in finance roles at Cisco and Microsoft, and has been the company's most familiar voice to Wall Street, appearing alongside Huang on nearly every earnings call of the past decade.

The board also appointed Harvey C. Jones as its first-ever chair. Jones has spent more than 30 years as an entrepreneur, technology executive and venture investor, including a stint as CEO and executive chairman of Synopsys. Nvidia had never had a board chair since its founding in 1993; creating the role now is a governance patch for the vacuum Huang leaves.

A search for a permanent CEO has started, with both internal and external candidates in scope. No timeline was given, which is itself the open question: how long does "interim" last?

The market's first read

Shares fell as much as 13% in premarket trading and closed down about 4.3%, wiping out roughly $182 billion in market value on the day. Measured from late October, when the first reports appeared, the drawdown is considerably larger; one widely cited figure puts the cumulative loss near $1.4 trillion.

AMD gained about 21% over five sessions. Several banks cut their price targets. Wall Street had been almost uniformly bullish on Nvidia for four years, so the shift in ratings matters as much as the price move.

How it got here

Reports in late October described a personal relationship between Huang and a former executive at a Singapore-based data center company. On November 8, Nvidia published a letter from Huang to employees in which he acknowledged infidelity, apologized to his family, staff, board, partners and shareholders, and wrote: "Leading NVIDIA is the privilege of my life. My responsibility is to serve you and be worthy of your trust. I have fallen short of that responsibility."

The board said the same day that an independent committee of directors, with outside counsel, would review "all available information," and stated it had no reason to believe national security or company trade secrets had been compromised. Three days later, the New York Times reported a second relationship involving an Nvidia executive; on November 12, he resigned.

Employees versus investors

Wired reported that before the resignation, a senior engineer posted on the company's internal suggestion board urging the board to consult staff first. The thread drew more than 1,600 comments, most supporting his return, and a staff-initiated poll showed over 97% wanted him to stay.

Bernstein framed the other side on November 9 with the term "key person risk": Huang drove long-term strategy and major operating decisions, reportedly had about 60 direct reports, and the management system depended heavily on him, with no clear successor.

Employees were judging his product instinct. Investors were judging the governance structure. Both are legitimate, and they answer different questions: can this team still build great chips, and can it do so without him.

Three things to watch

First, earnings on November 17. Nvidia confirmed its fiscal 2027 third-quarter results will be released as scheduled. Beyond the numbers, Kress's commentary on the next product cycle, supply commitments and capital spending will carry the signal.

Second, who gets the permanent job. An internal pick means continuity; an external one means recalibration. Either choice sets customer confidence for the 2027-2028 roadmap.

Third, whether orders move. Chip procurement is a long-cycle decision driven by performance, ecosystem and supply guarantees, so reputational damage usually transmits slowly. A leadership vacuum that slows delivery is a different problem, and it transmits fast.

The durable moat was never one person. It is CUDA's switching cost and a yearly cadence of new silicon. Neither disappears because a CEO leaves, but both decay if decisions start taking longer.