Jensen Huang Resigns as NVIDIA CEO; Multiple A-Share Supply Chain Companies Say Cooperation Unaffected
NVIDIA founder Jensen Huang has resigned from his positions as president, chief executive officer, and board member, effective immediately, with Colette Kress listed as interim chief executive officer. On November 13, share prices of NVIDIA supply chain companies on the A-share market fell across the board: TFC Optical Communication and Victory Giant Technology dropped about 10%, Shengyi Technology hit the limit down, and Zhongji Innolight fell about 8%. However, multiple suppliers said their cooperation with NVIDIA remains unchanged, and computing power vendors said prices for NVIDIA computing power cards have not fluctuated significantly. Analyst Gu Zhengshu believes that Huang's departure may affect the pace of NVIDIA's advancement in areas such as CPO and silicon photonic engines, while industry observer Liu Chao expects that the open-source large model ecosystem may suffer a setback; during the same period, AMD, Broadcom, and domestic computing power stocks such as Enflame Technology and Cambricon strengthened.
Why read this
SelectedAI score 78
The article summarizes responses from multiple parties—A-share supply chain companies, domestic computing power vendors, and analysts—regarding Jensen Huang's resignation, offering a view of the actual reactions across various links in the industry chain.
Topics
Body · AI translation
Article body
From issuing an apology statement to resigning, Nvidia founder Jensen Huang stepped down within a week from the chip empire he had run for 33 years, triggering turmoil in the capital markets and concerns across the AI industry chain. Over the past two days, Nvidia's stock price has fallen continuously, and the company's rating has been downgraded. On November 13, the stock prices of Nvidia's A-share industry chain partners fell across the board.
However, Nvidia's A-share partners said in their latest responses that there has been no change in their business cooperation with Nvidia. In addition, domestic computing power manufacturers also told reporters that the prices of Nvidia computing power cards have not fluctuated significantly. Industry insiders said that Huang's resignation is likely to affect the global AI industry chain and may also deal a blow to the open-source model ecosystem, and that any wavering in Nvidia's ecosystem strategy or any lag during the transition period could create a favorable window of opportunity for competitors.
A-share supply chain says "unaffected"
On November 13, Nvidia's official website updated its executive profile page, listing Colette Kress as interim chief executive officer while continuing to serve as executive vice president and chief financial officer, while Huang is listed as a member of the executive team and only as founder. A November 12 official Nvidia notice showed that Huang had resigned from the positions of president, chief executive officer and board member, effective immediately.
On November 13, A-share companies in Nvidia's industry chain were hit, with TFC Optical Communication and Victory Giant Technology falling about 10%, Shengyi Technology hitting the limit down, and Zhongji Innolight dropping about 8%.
Regarding the impact of Huang's resignation, a domestic listed supplier to Nvidia told reporters that it had not noticed any change in cooperation so far. In addition, several computing power manufacturers also told reporters that the current prices of Nvidia computing power cards have not shown significant fluctuations.
RunJian Co., Ltd. responded on an interactive platform to a question about whether it had been affected by the recent Nvidia incident, saying that under the token factory business model, the supply side focuses on token output, and the company also improves token output through technical means such as computing power and model optimization. The allocation of computing power resources is more flexible, and the types of servers supporting the computing power business are also more diverse. The company's procurement channels are diversified, which can guarantee computing power demand.
According to industry information, Nvidia's Vera Rubin series racks using a liquid cooling architecture will enter mass production and shipment in the fourth quarter. Regarding the latest progress, an executive of Lingyi iTech pointed out during a research survey that the company's Liminda, as a core supplier to a leading North American computing power customer, continues to deliver server liquid cooling and power-related products in batches to leading overseas computing power companies and their supply chain partners, related server contract manufacturers, power solution companies and other customers. Recently, Liminda, as an ecosystem partner of a North American computing power customer, exhibited core liquid cooling products such as the InnerManifold manifold and the new-generation ZQD (zero pressure drop quick connector).
In addition, recent online rumors claimed that due to quality problems with M9 copper-clad laminates, Shengyi Technology had its orders canceled by Nvidia. In response, Shengyi Technology said it had not received any notice of order cancellation from any end customer, and that its current production and operations are normal and orders are full.
Siemens CEO Roland Busch also said in a media interview that Siemens has joint projects with Nvidia and that he has not seen any impact on their cooperation so far. The outside world is focusing excessively on a single individual at Nvidia; regardless of what decision Nvidia's board makes on management issues, the company itself will continue to operate.
Since Huang's personal extramarital affair scandal intensified in late October, Nvidia's market value has fallen by $1 trillion. Recently, a number of Wall Street institutions have lowered Nvidia's target price and rating, while some institutions have maintained their ratings, emphasizing that Nvidia's "supply chain moat remains solid."
Regarding the impact of Huang's resignation, Gu Zhengshu, a semiconductor industry analyst at Moyuan Think Tank, said that as the soul figure of Nvidia and the chief architect of its full-stack accelerated computing system, Huang's departure is likely to bring far-reaching chain reactions to the global AI chip and advanced packaging industry chain. This is because Huang led Nvidia's transformation from a single GPU chip to a full-stack system-level approach. If leadership changes, Nvidia's extremely aggressive pace of iteration in software and hardware may face fine-tuning in organizational management, commercialization and R&D cost balance.
Advanced packaging and the core supply chain may also be affected. Among them, the variables in advancing CPO (co-packaged optics) and next-generation interconnect routes — Nvidia's route choices in co-packaged optics and silicon photonic engines — are a "weather vane" for the global industry. Whether a change in company leadership will affect its aggressive bet on all-optical interconnect routes will directly affect R&D expectations across the upstream optical module and optoelectronic integration industry chain.
Beyond hardware, Huang's efforts at the financing level to build Nvidia into an "AI central bank" may also change. According to incomplete media statistics, Nvidia's related financial commitments amount to about $164.5 billion, involving Open AI data centers, leasing arrangements and other capacity guarantee projects, which have already raised investor concerns about the company's efficiency and safety in using funds, while Colette Kress had previously taken a conservative stance.
In addition, industry observer Liu Chao expects that open-source large models may suffer a setback because of the Huang incident. Huang once clearly supported the open-source ecosystem, repeatedly praising "Chinese models as excellent," opposing the use of "security" as an excuse to ban open-weight models across the board, and even establishing the "Open Secure AI Alliance."
On November 13, Zhipu's stock price fell 10%, and MINIMAX-W fell about 8%.
Domestic computing power accelerates its breakthrough
Recently, Nvidia's competitors have continued to rise. AMD's stock price has risen about 30% cumulatively this week, and Broadcom's stock price has risen about 24% cumulatively. In addition, domestic computing power chip stocks have strengthened: this week, Enflame Technology surged nearly 55%, Cambricon rose nearly 22% cumulatively, and MetaX and Moore Threads rose nearly 16% and 10% cumulatively, respectively.
Regarding the impact of Jensen Huang's resignation on domestic computing power, Gu Zhengshu analyzed that the biggest competitive barrier for the current local computing power ecosystem lies in Nvidia's CUDA monopoly and its closed loop of software and hardware ecosystem. If Nvidia's ecosystem strategy shows any wavering or lags during a transition period, this will objectively provide a more ample window for the local ecosystem in terms of framework compatibility and the implementation of heterogeneous mixed-training platforms.
From a performance perspective, domestic computing power chip stocks have generally seen profit growth: Cambricon achieved net profit of approximately 3.7 billion yuan in the first three quarters of this year, a year-on-year increase of 1.31 times; MetaX's profit grew 2.85 times year-on-year; and Moore Threads turned a profit year-on-year.
The competitive landscape of the domestic computing power market is also changing.
Moore Threads executives said in a recent institutional research meeting that market competition will shift from simply supplying goods to competing on product value, including computing power, efficiency, and algorithm collaboration. The company adheres to the full-function GPU route and software compatibility; its ecosystem has already covered 800,000 developers. The company's customers include internet and cloud service providers, token service providers, AI innovation enterprises, telecom operators and intelligent computing centers, research institutes, government and enterprise units, and industry customers, with scenario requirements covering model training, inference, and physical AI, among other directions.
On the client side, Chinese cloud vendors, telecom operators, and others are increasing capital expenditure. At a recent mobile network conference, China Mobile said it plans to further increase total procurement in 2027, boost investment in computing power, and ensure a doubling of intelligent computing capacity. In the next two years, its intelligent computing capacity will double consecutively, and it will build multiple gigawatt-level intelligent computing parks.
According to a Morgan Stanley report, China's AI computing power industry is expected to enter a new stage driven jointly by a substantial expansion of supply from the domestic chip supply chain, capital investment in AI cloud computing, and a wave of AI inference commercialization. It is estimated that from 2026 to 2030, hyperscale cloud computing vendors, emerging AI computing power cloud enterprises, major telecom operators, and other emerging AI entities will invest a cumulative total of about 8.5 trillion yuan, of which 9.95 trillion yuan will be reached in 2026, a year-on-year increase of 121%.
(Source: Securities Times) Mid-article embedded advertisement