Jensen Huang Steps Down as NVIDIA CEO, CFO Kress Takes Over Temporarily
On November 12, Jensen Huang resigned from his positions as NVIDIA's President, CEO, and board member, ending 33 years at the helm, and Executive Vice President and CFO Colette Kress took over as interim CEO. The article notes that NVIDIA's product roadmap has been planned through 2028 and the NVLink Fusion ecosystem is still expanding, but with Microsoft's Maia 200 going into production and Amazon's Trainium2 handling most of Bedrock's inference, major customers are accelerating their own chip development. NVIDIA has also partnered with institutions including BlackRock, Blackstone, and Goldman Sachs to build a compute financing platform, planning to leverage $500 billion in third-party capital to support customers' GPU purchases, and according to Reuters, Kress had previously raised objections to such high-risk investment arrangements.
Why read this
AI score 59
The article outlines the threefold pressure NVIDIA faces after Jensen Huang's departure, including major customers developing their own chips and the risks of the compute financing model.
Topics
Body · AI translation
**After Jensen Huang steps down, Nvidia will still hold its technology roadmap and NVLink ecosystem moat in the short term, but with major customers developing their own chips, the AI bubble, and "AI central bank"-style financing risks piling up, whether his successor can continue the aggressive investment strategy becomes an even bigger test.**
**Key points**
**1. Jensen Huang steps down, CFO temporarily takes over**
On November 12, Jensen Huang resigned as president, CEO and director, ending 33 years at the helm; Executive Vice President and CFO Colette Kress became interim CEO.
**2. Short-term technology moat remains solid**
The product roadmap is already planned through 2028, and the NVLink, NVSwitch and NVLink Fusion ecosystems form a complete AI factory platform; the advantage will not collapse immediately, but how long it can last is questionable.
**3. Major customers' in-house chips accelerate "de-Nvidia-ization"**
Microsoft's Maia 200 has entered production, Amazon's Trainium2 handles most Bedrock inference, and OpenAI, Anthropic and others are also developing their own chips; Anthropic disclosed that a model can be ported to AMD MI355X in about 3 days.
**4. The "AI central bank" model depends on real demand**
Nvidia uses compute capacity leasebacks, long-term purchase commitments and partnerships with asset managers to leverage $500 billion in capital to support customers buying GPUs; if real AI demand cannot keep up, the risk will spread to the industry and the financial system.
**5. Whether the aggressive investment strategy continues is uncertain**
Jensen Huang tended to support the entire AI ecosystem rather than betting only on winners. But according to Reuters, former CFO Kress had opposed high-risk investments and financing arrangements, and whether the successor will continue them is uncertain. 2026-11-14 09:41 Nvidia has bigger troubles than Jensen Huang's departure NoNoise Quick Take
This article is from the WeChat public account: NoNoise, editor: Xiaowei, author: Liu Shiyu
The Jensen Huang era at Nvidia has ended.
On November 12, Jensen Huang resigned as Nvidia's president, chief executive officer and director, ending a 33-year tenure at the helm. Executive Vice President and Chief Financial Officer Colette Kress became interim CEO until a formal successor is determined.
Since Nvidia was founded in 1993, Jensen Huang has been almost synonymous with the company. Nvidia long adopted a highly flat management structure, with about 60 subordinates reporting directly to Jensen Huang, and the company had no recognized No. 2 figure. Every strategic decision by Jensen Huang would set off chain reactions across the entire AI industry.
This sudden dramatic personnel change has brought Nvidia an unprecedented test.
For now, Jensen Huang has left Nvidia a moat that is difficult to surmount in the short term, but that is not enough to solve all the problems the company faces.
01
"Something completely different"
According to Vanda Analytics data, on the day Jensen Huang announced his resignation, retail investors bought about $111 million of Nvidia stock within half an hour after the U.S. market opened, the largest opening-period retail buying on record for the stock.
Retail investors' confidence is not hard to understand: "Be greedy when others are fearful." Jensen Huang is leaving, but Nvidia's technology and ecosystem advantages remain.
First, Nvidia's technology roadmap for the next few years has long been clearly planned.
As early as March 2025, Nvidia publicly disclosed a product roadmap extending to 2028. From Blackwell and Rubin to Feynman, the iteration directions of core technologies such as GPUs and interconnect networks have all been clearly arranged.
▲The roadmap shown in Jensen Huang's GTC2025 keynote
This means that, at least in the short term, Nvidia can still advance product R&D and iteration along the established technology roadmap.
Moreover, Nvidia's moat is not just leadership in chip performance.
On Nvidia's earnings call on August 26 this year, an analyst pressed Jensen Huang: OpenAI and Anthropic are accepting NVIDIA's investment and support while developing their own AI chips; how does Nvidia view this competition?
Jensen Huang's answer was: "We're building something very different."
He also said he was "100 percent confident" that OpenAI, Anthropic and other companies will continue to use Nvidia's computing platform for a long time to come.
A partnership in August this year explains Jensen Huang's confidence. MediaTek reached a cooperation with Nvidia to use the latter's NVLink Fusion technology to develop custom AI chips for customers and connect them to Nvidia's data center systems. At the same time, Nvidia spent $3.5 billion to buy convertible bonds issued by MediaTek.
At first glance, this partnership is rather subtle. MediaTek is aggressively entering the custom AI chip market and originally came to compete for Nvidia's GPU business, so why did it instead choose cooperation?
The reason is that building a powerful chip and getting thousands of chips to work together efficiently are two different things.
Large AI data centers not only need chips, but also need to enable high-speed data exchange between chips and connect large numbers of computing devices into huge clusters. The industry calls these two processes Scale-up and Scale-out, respectively.
For AI data centers, what truly matters is not just the performance of a single chip, but how many computing tasks can be completed and how many tokens can be generated for the same money and the same amount of electricity.
This is precisely where Nvidia's advantage lies. It uses NVLink and NVSwitch to achieve high-speed interconnection between chips, and then uses networking technologies such as InfiniBand and Spectrum-X to organize computing devices into large-scale AI clusters.
NVLink Fusion, launched in 2025, further opened up this capability to other chip companies. In addition to MediaTek, companies such as Amazon, Fujitsu, Qualcomm, Arm, and Marvell have also joined this cooperative ecosystem.
▲NVLink Fusion ecosystem partner board
Nvidia's calculation is that even if some GPU orders flow to custom chips in the future, as long as those chips still use Nvidia's interconnect and networking systems, it will have a chance to continue profiting from them.
English text is being prepared.
These competitive advantages formed over many years will not easily collapse with Jensen Huang's departure, but how long they can be maintained in the future is hard to say. Because every brick in this barrier bears the distinct imprint of Jensen Huang himself.
Having lost its helmsman, Nvidia now faces a situation more complex than ever before.
Nvidia's major customers are precisely the ones who most want to reduce their dependence on it.
Cloud giants such as Google, Amazon, and Microsoft, which purchase large quantities of GPUs from Nvidia every year, are accelerating their own ASIC (application-specific chip) development; AI model companies such as OpenAI and Anthropic have also begun developing their own chips, attempting to control everything from models all the way to computing power.
▲Compiled and charted based on publicly available online information: NoNoise
In October this year, Microsoft announced that its self-developed chip Maia 200 had officially entered production and use, began supporting businesses such as Microsoft Copilot, and plans to expand deployment in 2027. Amazon is also accelerating deployment of its self-developed chip Trainium. According to data disclosed earlier this year, Trainium2 already handles most of the inference tasks on the Amazon Bedrock platform.
For a long time, Nvidia's chips have been in short supply, giving it strong bargaining power; even tech giants had to personally step forward to secure supply.
Oracle founder Larry Ellison once recalled that at a dinner, he and Musk repeatedly begged Jensen Huang to sell them more GPUs, and even kept saying they hoped Nvidia would charge more. In the end, he added that the dinner did indeed work.
The continuous iteration of AI technology has already shown the industry a glimmer of possibility for challenging the moat of the CUDA ecosystem.
In November this year, Anthropic disclosed a technological advance during its IPO roadshow. In the past, when the company connected to a new chip platform, it often took dozens of engineers several months to port and optimize the underlying computing programs.
But with the help of its self-developed Claude Mythos Preview model, the team needed only about 3 days to bring a model's prompt-processing task to usable performance on AMD MI355X chips, with almost all related computing kernels ported or written by AI.
▲In April this year, Google and Anthropic reached a partnership
to integrate Claude Mythos into Google Cloud's agent development foundation
Compared with model training, inference tasks have more diverse scenarios. Although Nvidia GPUs have the advantage of generality, for large-scale inference tasks with relatively stable demand, specialized chips can be optimized in a targeted way to complete computing at lower cost.
This means that major customers have both the opportunity and the incentive to shift part of their computing demand to self-built platforms, gradually reducing their dependence on Nvidia.
03
Therefore, Nvidia not only sells shovels, but also needs to ensure that the miners have money to buy shovels.
The first step is to personally provide a backstop for customers.
Small and medium-sized cloud providers want to seize the AI infrastructure business, but do not dare to bet all their assets on GPUs. Equipment is easy to buy, but no one can guarantee whether the computing power can be rented out. Nvidia's proposal is: go ahead and purchase boldly and build data centers; if there is idle computing power in the future, I will lease it back or buy it according to the agreement.
CoreWeave is a typical case. Facing market doubts about computing power oversupply and high expenses, Nvidia offered a long-term computing power purchase commitment, equivalent to adding a layer of insurance to its future revenue. Cloud providers therefore find it easier to raise financing, get money to continue buying GPUs, and Nvidia also stabilizes orders.
But if the entire industry relies on Nvidia to provide a backstop, even the richest cash flow could not sustain it. So the second step came: bring Wall Street on board as well.
On August 10, 2026, Nvidia's official website formally announced the establishment of a computing power financing platform together with six major Wall Street asset management institutions including BlackRock, Blackstone, and Goldman Sachs, planning to leverage $500 billion in third-party capital to lend to AI companies/cloud providers to purchase GPUs and build AI data centers.
English text is being prepared.
Moreover, the operation of this "computing power central bank" cannot be separated from Jensen Huang's own strategic will.
In April this year, when talking about his investment logic on a podcast, he said that Nvidia does not intend to pick only winners, but hopes to support everyone. This philosophy is not hard to understand. For Nvidia, as long as the entire AI industry continues to develop, no matter which company ultimately wins, it has a chance to become a customer of its chips and computing power platform.
What Jensen Huang values is not only the financial return of a single investment, but also the long-term benefits brought by the growth of the entire industry.
But this aggressive investment style is not without controversy inside Nvidia.
According to Reuters, current interim CEO and former Chief Financial Officer Colette Kress has repeatedly raised objections to the high-risk investments and financing arrangements promoted by Jensen Huang, and the disagreements between the two involved companies such as OpenAI and AI infrastructure company Lambda.
In the past, Jensen Huang, by virtue of his founder status and long-accumulated authority, was able to play a leading role in major investments. Now, does his successor still have enough boldness to continue this aggressive investment strategy?
Has published 70 articles on Huxiu. This content comes from the internet, original link; the views represent only the author and do not represent Huxiu's position. If there is any copyright issue, please contact hezuo@huxiu.com, and we will verify and handle it promptly. Everyone who is changing and wants to change the world is on the Huxiu APP[ ]