Jensen Huang has left Nvidia. The infrastructure built around Nvidia is going to take considerably longer to move.
That distinction matters as the company confronts an extraordinary leadership crisis. Huang helped build Nvidia into a central supplier to the AI industry, and replacing him creates uncertainty that extends well beyond the corner office.
My take: Nvidia will probably be fine over the long run. Getting there could be rough.
That is a judgment about the durability of the business. It doesn't excuse Huang’s admitted conduct, settle the allegations surrounding him, or tell us whether the board handled his departure correctly. Those questions deserve their own scrutiny.
On November 12, Nvidia announced that Huang had resigned as president, CEO, and a board member, effective immediately. Chief Financial Officer Colette Kress becomes interim CEO while retaining her finance responsibilities. The company is searching for a permanent successor and will consider internal and external candidates.
Harvey C. Jones, a former Synopsys CEO and executive chairman, also becomes Nvidia’s first board chair.
Those are the confirmed leadership changes. The circumstances surrounding them require a little more care.
Huang acknowledged being unfaithful to his wife in a November 8 message to employees. Separately, Nvidia’s board opened an investigation into reporting that he had been targeted by a Chinese intelligence effort. At the time, the board said it had no reason to believe national security or company secrets had been compromised.
That was the board’s position when it announced the inquiry. It was not a completed investigative finding.
On November 11, The New York Times reported that Huang had also had an extramarital relationship with Kimberly Powell, Nvidia’s healthcare vice president, who reported to him. Reuters relayed that reporting, attributing it to the Times and five people who had worked at Nvidia.
We should keep those distinctions intact: an admission of infidelity, a reported workplace relationship, and unresolved national security allegations. The resignation announcement does not resolve the investigation or disclose the board’s full reasoning.
Huang’s admitted conduct is unacceptable. Whether the board’s handling of the crisis proves sound will take more information and time to assess. A company’s eventual financial performance, by itself, won’t answer that question.
For now, Kress has the immediate job of keeping Nvidia moving.
Her background is substantial. She joined Nvidia as CFO in September 2013 after senior finance roles at Cisco, 13 years at Microsoft, and earlier work at Texas Instruments. At Microsoft, she served as CFO of the Server and Tools division.
She brings familiarity with Nvidia’s finances, operations, and business relationships. That makes continuity a reasonable expectation. Whether she becomes the permanent CEO is unknown.
The permanent search also needs more discipline than the usual internet exercise of naming every executive with a recognizable face.
Reuters’ succession reporting highlights Ian Buck, whose work helped develop CUDA and who leads Nvidia’s hyperscale AI products, as an internal executive favored by some analysts (and, full disclosure, so do I.) Arm CEO and former Nvidia executive Rene Haas has also been discussed as an external possibility.
Madison Huang and Tim Cook have appeared in public succession speculation. Neither is a confirmed candidate.
Madison, Huang’s daughter, oversees physical AI product marketing as a senior director, according to Reuters. That role gives her relevant experience; it does not establish that the board is considering her for CEO.
Cook is an interesting possibility on paper who could potentially make an excellent fit. His experience running a huge global technology business, coordinating manufacturing, and managing supplier relationships would be relevant. He also knows what it means to succeed a founder whose identity is inseparable from the company.
But Cook remains Apple’s executive chairman following its September CEO transition. There is no confirmation here that he wants the role or has even been approached. A plausible fit is several steps away from an actual candidacy.
Whoever takes over will inherit advantages that survive a leadership announcement. Also, the buffer of an interim CEO, saves them from immediately dealing with one of the largest leadership crises ever.
Nvidia’s business reaches across GPUs, networking, systems, and software. CUDA—the programming platform developers use to put Nvidia hardware to work—is especially important. The company’s 2026 annual report describes more than 7.5 million developers using CUDA and its other software tools.
Customers have built applications, trained teams, and organized infrastructure around that ecosystem. Our GTC 2026 breakdown traces how widely Nvidia’s ambitions now extend across AI computing.
Changing suppliers can mean rewriting or adapting software, validating performance, securing hardware, and changing how systems operate. At large scale, that becomes a substantial undertaking.
Capacity adds another constraint. During Nvidia’s August earnings call, Huang described a supply chain running at full stretch and infrastructure preparations that could extend two to three years.
Those comments came from Nvidia’s own management. They support the argument that compute capacity takes time to build; they do not prove every customer would need years to switch.
Some customers could redirect individual workloads or future orders much sooner. Nvidia itself acknowledges competition from rivals and customers building their own technology.
Still, an industry-wide departure would require far more than a desire to leave. The software, equipment, capacity, and operational work all have to line up. That is a major reason I expect Nvidia to remain central to AI.
The bigger danger may emerge gradually. Customers can continue buying Nvidia today while accelerating plans to depend on it less tomorrow. A leadership crisis could strengthen that incentive without producing an immediate collapse in orders.
And Nvidia still has to earn its future. Delivering an established roadmap is one challenge. Deciding where computing goes next is another. The next CEO will need to retain technical leaders, maintain customer and government trust, and make decisions Huang previously drove.
One direction worth watching closely is open-source AI.
Huang has been a prominent advocate for models that organizations and individuals can download, adapt, and run themselves. Nvidia’s July paper, Open Weights and American AI Leadership, argues that broader access supports competition, innovation, and greater control over AI.
That position has been accompanied by products. Nvidia’s Nemotron work, including the multimodal models, gives developers more ways to build their own systems. On November 10, the company also announced new open-source infrastructure contributions, including OpenShell.
There is a business reason to keep investing in that direction: more organizations building and running AI can create more demand for compute.
I expect Nvidia to continue supporting open models. That remains an expectation, not a commitment from its future permanent CEO. And “open weights” should not be treated as a guarantee of unrestricted use; individual licenses still matter.
For builders and businesses, the practical response is to watch execution: product delivery, support, leadership retention, licensing, and the pace of investment in the tools they rely on.
Nvidia’s November 17 earnings call is the next scheduled opportunity for management to answer questions. Over the months that follow, the company will have to demonstrate that its leadership team can sustain the work Huang helped set in motion.
My take: I think Nvidia will probably be fine long term. Its customers have built too much around its technology for that position to disappear overnight. What remains uncertain is how well the company manages the transition and how much ground competitors gain while it does.
Related reading
- Everything Nvidia Announced at GTC 2026, Explained — The hardware, software, and infrastructure strategy Nvidia’s next leadership team inherits.
- Nvidia’s Nemotron 3 Nano Omni Gives AI Agents Eyes and Ears — A closer look at the company’s investment in open models for developers.
- The Industry’s Push to Keep Open-Weight AI Open — Context on the broader debate over downloadable models, access, and competition.
