Nvidia's New Chair Last Ran a Company Before Nvidia Made a Chip
Nvidia's board was built for a founder who'd never leave. Now he's gone, and the board that has to pick his successor is short on recent operating experience.
For 33 years, Nvidia's board ran without a chair. Six months ago it was still explaining why. In its proxy statement, the filing every public company sends shareholders before its annual meeting, it said going without one "ensures that each member has an equal voice."
Then on November 11 the board accepted Jensen Huang's resignation as President, CEO, and director, effective immediately. The same day, it named its first-ever chair.
A board that didn't need a chair for three decades needed one the minute the founder walked out the door.
Huang's affair, the reports that he'd been targeted by a foreign intelligence effort, the board's investigation: that's been covered plenty, and no board anywhere can stop a CEO from making bad personal choices. What I can't stop thinking about is the board, and a question nobody has really asked yet.
Who in that room is ready to pick the next CEO of Nvidia?
Give them their due
Credit first, because they've earned a lot of it.
Some of these directors were there when Nvidia's first chip came out in 1995 and didn't take off, and they were there for the run that turned Nvidia into one of the most valuable companies in the world. Jensen set the strategy, the board backed him, and it worked about as well as anything has ever worked. A founder gets room a hired CEO doesn't, and this board gave him plenty.
When things blew up this month, the board moved. It opened an investigation on November 8, and three days later the CEO was out. I've sat on a public board that removed its CEO. For us it wasn't a scandal, it was a change in direction, and it still wasn't easy.
So maybe this board was fine. Maybe it did exactly what a good board does.
Except for one thing.
On November 8, the same day it opened the investigation, the board put out a statement. "The Board has confidence in Jensen and his leadership team."
Three days later he was gone.
Changing your mind in three days isn't a crime. Things move fast in a crisis. But it made me want to look at who was actually sitting in that room, so I did. I read Nvidia's proxy, plus nine others: five big tech companies and four chipmakers. It's the same document every shareholder gets before the annual shareholder meeting.
Who's in the room
One statistic that stood out when I was looking at the proxy documents was the longevity of the board members.
Nvidia's outside directors average 14.8 years. The S&P 500 average is 7.9 years, according to Spencer Stuart, the executive search firm that tracks board data.
Tench Coxe and Harvey Jones joined in 1993, the year Nvidia started. Nobody on any of the other nine boards has been there that long as an outside director. Coxe came from Sutter Hill. Mark Stevens, who's been on and off the board since 1993, came from Sequoia. Those are the two venture firms that backed Nvidia at the start.
That's how startup boards usually get built. When we took Teligent, the telecom company I co-founded, public, our board was mostly investors. When somebody sold a big stake, they left, and whoever bought it took their seat. The board changed hands. It didn't change shape.
Brooke Seawell joined Nvidia's board in 1997. He and Jones ran Synopsys, the chip-design software company, together in the early '90s, Jones as CEO and Seawell as CFO. The lead director, Stephen Neal, used to run Cooley, the law firm that formed Nvidia in 1993 and has been its main outside counsel for decades.
And here's the part most people miss. This board picks its members. At Nvidia, the committee that recommends directors is chaired by Neal, while Jones and Stevens sit on it too. Every year that committee decided who went on the ballot, and every year the long-timers went back on. Nvidia's own governance policies say term limits and mandatory retirement ages "are not in the best interests of the Company."
None of that breaks a rule. They all count as independent. The question is why the board still looked like that at Nvidia's size.
Then I asked a different question, director by director. When did this person last actually run something? A company, a business unit or a function. Board seats, law, consulting, investing, academia and government don't count as operating experience.
Here's what came back:
Studio Notes · Phil McKinney
| Board | Median years since last operating role | Share with operating role in last 10 years |
|---|---|---|
| Nvidia | 24 | 30% |
| AMD | 11 | 43% |
| Broadcom | 10 | 57% |
| Alphabet | 7 | 56% |
| Intel | 7 | 70% |
| Amazon | 5 | 70% |
| Qualcomm | 5 | 80% |
| Apple | 4 | 71% |
| Microsoft | 2 | 80% |
| Meta | 1 | 55% |
Operating role: ran a company or a business function. Sitting CEOs excluded. No operating role counts as the longest gap. Medians rounded to the nearest year.
Source: 2026 proxy statements.
© 2026 The Innovators Studio, LLC · Share freely with credit
Download PNG for sharing
At least half of Nvidia's directors are 20 or more years from running anything, or never have. Only three have run anything in the last ten years. Dawn Hudson was CMO of the NFL until 2018. Melissa Lora ran Taco Bell International until 2018. Aarti Shah was CIO at Eli Lilly until 2021.
You can usually tell when a director's experience has gone stale. They'll ask some clarifying questions, but they don't push back. The boards I've seen push hardest were full of people who were running something right now, or had been until recently, and who'd been recruited for exactly that.
The chair
Which brings me back to Harvey Jones.
Jones co-founded Daisy Systems in 1981 and became CEO of Synopsys in 1987. In January 1994 he handed that job to Aart de Geus. Since 2004 he's run his family office.
Nvidia announced its first chip in May 1995.
The man now leading the board that has to pick Nvidia's next CEO stopped running a company more than a year before Nvidia had its first product.
It doesn't get much better on technology. At Nvidia, the strategy is the technology, so I checked whether anyone on the board has ever led engineering, product or research in tech.
Nobody has. Zero of ten.
Jones has a computer science degree, but he came up through sales and marketing. Stevens has engineering degrees and did early-career technical work at Hughes and technical sales at Intel before he went into venture capital. Shah ran IT at a drug company. John Dabiri is a serious scientist at Caltech, but his field is fluid dynamics. Nvidia's proxy does credit Jones and Stevens with "Industry & Technical" skills, and I'd expect the company to make that argument. My bar is narrower. Has anyone in the room actually led technical work?
I've stood in front of a board committee like that. At HP, technology deals over $200 million went to the board's technology committee. My role was to present M&A deals for approval, including Palm, the smartphone maker, a $1.2 billion deal in April 2010. There weren't many questions. The people on that committee were mostly finance and business executives who had never held a technical role, and the two who had were years removed from it. What came back was caution. Nobody was trying to make the strategy better. They were trying to make sure nothing went wrong.
To be fair, Apple's and Microsoft's boards are also at zero on technical leadership, so not every tech board needs engineers. But their directors have run things recently, with medians of 4 years and 2 years. Nvidia is the only one of the ten that's thin on technical depth and recent operating experience at the same time.
The board leaned on Jensen for both. And now he's gone.
I've seen this movie
This is the part that worries me, because I've watched a board try to pick a CEO it wasn't equipped to judge.
When Mark Hurd left HP in a scandal in 2010, the board went outside and hired Léo Apotheker from SAP. SAP was a fraction of HP's size, and he'd never run anything close to HP's scale. I spent a couple of hours with him in his first month and walked out of that room shaking my head, not sure how the board had picked him. He lasted about 11 months. Fortune later called it "an impaired board and an ill-chosen CEO."
That's the call Nvidia's board has to make now, with less recent operating and technical experience in the room than HP had.
The job nobody did
If there's one place this board owns what happened, it's succession planning.
Nvidia knew the risk. Its own 10-K warns that failing to ensure effective succession planning could hinder the company's strategic planning, execution and long-term success.
Now look at the proxy. Every one of the other nine boards I checked says something about how it handles CEO succession. Nvidia's proxy doesn't have a section on it at all. The closest it gets is one line about believing in "extensive leadership development."
And Huang wasn't shy about where he stood. At Stanford in 2009 he called picking out successors ahead of time "a toxic, toxic process." This March, on Lex Fridman's podcast, he said it straight: "I'm famous in saying that I don't believe in succession planning."
When a CEO says that, the board should sound the alarm. Nobody, in any role, is indispensable. This board let it stand.
Naming the CFO as interim CEO within hours is what an emergency plan looks like, and I'll give them credit for that. What was missing was the real plan. There was no obvious number two. Dozens of people reported straight to him. If anyone was being prepared to take over, nobody outside the company could see it.
Here's what a real plan looks like. For every senior role, not just the CEO, you name three groups of people:
- Who could step in right away if that person got hit by a bus. That's your bench strength, and it's what keeps the company standing on a bad day.
- Who could be ready in about three years. They get rotated between roles, given special projects and stretched on purpose.
- Who might be ready in five years or more. That's your early-career talent pool, and you invest in them now.
The best boards I've seen set aside one meeting every year for this. The CEO and the chief people officer walk the board through the succession plan for every senior role. The people on that plan come into board meetings throughout the year and present, so the board gets to see the bench of talent in action.
What Nvidia's board should do now
If I were in that room this week, I'd start with the board, not the CEO search.
The board needs to refresh itself, and it needs a technical bench of its own. Bring on people with recent operating experience and recent senior technical leadership before the CEO decision, not after it. Those are the people who can tell a good candidate from a good interview.
Then build a real succession plan, for the whole senior team, and put it in front of the board every year.
And deal with tenure. A board seat is about the closest thing to guaranteed employment in business, and getting a long-serving director to step off is really hard. A set term that can be extended twice, and then you rotate off, takes the awkward conversation off the table.
Nvidia's 2027 shareholder meeting, and the director election that comes with it, will be one to watch.
Your board
Shareholders don't get a pass here either. Nvidia's two biggest holders are BlackRock and Vanguard, each with a little over 7%. Huang said in March that he doesn't believe in succession planning. In June, every director was re-elected with at least 86% of the vote. Where were BlackRock and Vanguard?
Every shareholder deserves a board that understands and practices good governance. It's table stakes, and every public company owes it to every shareholder, not just the big ones. The trouble is that nobody agrees on what "good governance" means, so it turns into hand-waving. So make it concrete. Pull up your own board's proxy and ask:
- Who on this board has done the CEO's job, at our scale, in the last ten years?
- Who could test our technical strategy, and not just approve it?
- When did our longest-serving director last vote against the CEO?
- If the CEO walked out tonight, who runs the company tomorrow? And who runs it two years from now?
If you can't answer that last one, you've got the same problem Nvidia has. You just haven't found out yet.
This article is for informational purposes only and does not constitute financial or investment advice or a recommendation to buy, sell, or hold any security. I may or may not own shares in companies mentioned, and my holdings may change without notice. Do your own research and consult a qualified financial adviser before making investment decisions.