Bloomberg Gets Its 'China Must Lose' Calculations Wrong Again
No sooner had the news of Jensen Huang's resignation broken than Bloomberg published a column titled, "China Is the Biggest Loser From Jensen Huang's Downfall." The logic can be summarized as follows: Huang was the one advocating for the sale of chips to China; once he leaves, restrictions will only tighten, China will be unable to buy Nvidia chips, and therefore China will be lost. Each link in this chain of reasoning appears to touch on some factual element, but when combined, the accounting leads to an incorrect conclusion.
Billing China for Washington's infighting was the first accounting error. The Chinese market did not drive Huang out. His resignation was prompted by an extramarital affair scandal and board investigation, but Congress pushed the affair toward "getting tough on China." That day, the chairman of the House committee focused on China called his departure a "firing" and demanded that the new leadership put national security before business. Huang had already agreed to testify before Congress on chip exports, and both parties were stepping up efforts to restrict chip sales to China. The trigger was pulled in Washington, not Beijing. Bloomberg called America's political reshuffle a "Chinese strategic failure," charging one household's bills to an outsider.
The second accounting error: using a ruler that measures almost nothing to gauge China's position.
Bloomberg's premise is that "China can still buy Nvidia chips," so once Huang leaves, China will no longer be able to buy them, and therefore it has lost. This premise is nearly unsustainable. On an earnings call, Nvidia's CFO stated that data center chips sold to China constituted a negligible portion of its total data center revenue, and the following quarter's forecast assumed no revenue from this segment. The company's revenue from China dropped from nearly 30 percent three years earlier to single digits. Later, Huang admitted that Nvidia's direct sales share of China's AI accelerator market had reached zero. Discussing "how painful losing it would be" with a depleted account leaves little to calculate.
The trend that really deserves attention is on the other side of Nvidia.
Industry research shows that Nvidia's share of China's data center AI chip market has fallen significantly, while Huawei's Ascend share has risen significantly, narrowing the gap between the two to single-digit percentage points. Chipmaker Cambricon's financial statements also show that its revenue multiplied several times over and that it recorded its first full-year profit since listing. Chinese-made chips remain far from catching up with Nvidia – manufacturing yields, supplies of high-bandwidth memory, and the software ecosystem are all hurdles that remain. However, they have already surpassed the question of "whether they exist at all." In declaring China a loser because it "cannot buy Nvidia," Bloomberg has missed precisely how this stage of the contest has changed.
Consider this industrial contest from a different perspective: for China's industry, "still being able to buy" has never served as a bargaining chip but rather as a brake pad. As long as the window for compliant procurement remains open, companies, concerned about migration costs and engineering risks, will default to an off-the-shelf solution with a mature ecosystem, delaying the moment when Chinese-made chips receive large-scale orders and feedback from real-world applications. Once that window narrows, procurement decisions are forced toward domestic alternatives, accelerating validation and iteration instead. Bloomberg reads "the brake pad has been removed" as "China has lost its steering wheel," reversing cause and effect.
Bloomberg did not listen carefully to Huang's own words.
On a podcast, he called the case for US export controls "extremely foolish" and a classic example of a "loser mentality." He was not speaking up for China; he was calculating an industrial balance sheet for America: artificially creating two separate ecosystems, an open-source system running only on a foreign technology stack and a closed system running on the American technology stack, "would be a very unfavorable outcome for the United States." Openness means life; closure means death – these words were not spoken on China's behalf, but on behalf of America's technological leadership.
America has heard this historical rhyme before.
In the 1980s, the US Senate overwhelmingly passed provisions imposing sanctions on Toshiba, barring its products from being exported to the United States for several years; Toshiba's chairman and president subsequently resigned. Alstom was hit with a substantial fine by the US Justice Department, and its core power business was eventually acquired by General Electric. The lesson is not that "suppression always fails," but that its costs are never borne solely by those being suppressed. Toshiba fell, but America did not regain its dominance in memory chips; Alstom was broken up, and what America lost was its European allies' trust in "American rules." It is easy to change a company's fate through suppression; changing an industry's geographical distribution and the direction in which technology spreads is challenging
The main issue with the "China must lose" narrative is that it's addictive, not that it's wrong. As long as a "loser" can be manufactured, there is no need to confront the troubles within one's own system. But the costs are real: when your analysis assumes that your opponent is always losing, you lose the ability to understand why that opponent has not lost. A system that needs to keep declaring that others have lost is often doing so because it is no longer quite sure whether it can still win.
On the day Huang resigned, Nvidia shares plunged at the open, pared their losses by the close, and posted a decline for the week. The market was worried not about whether China would buy chips, but about whether Nvidia, after losing its defining figure, could still preserve the technology ecosystem built on a foundation of openness.
What truly warrants concern is not how China will respond, but whether, after America has driven out someone like Huang, anyone will still dare to say that "openness means life." Whether he stays or leaves is America's own affair. But if the approach he represents – that technology ecosystems should be open and the pool of developers should be as large as possible – becomes something no one in America dares to advocate publicly after his departure, then America will have lost more than a CEO. It will have lost the most valuable part of the confidence underpinning its technological leadership.
(The author is a former official with the Information Office of Shanghai Municipality. The story, originally in Chinese, was translated by Shanghai Daily. The views are his own.)
Editor: Wang Xiang
In Case You Missed It...

![[ChinaMaxxing] The Weather We're Having Now Is Called Xiaoxue](https://obj.shine.cn/files/2026/11/20/29057db6-0ed4-4e29-acb7-228e358d27fa_0.png)
![[ChinaMaxxing] The Weather We're Having Now Is Called Xiaoxue](https://xinyue-v30o.yueliu.cloud/coverImg/project/video/c035e2a5b858931f21ef098ec87016c7-0a6b719283e54b7e33694edea340626c/j56598e0ce4b011f1aa66850ce592694f_0.jpg)

Popular Reads

Global Tech Selloff Snares Chinese Stocks Amid Doubts About the AI Boom

Shanghai, Singapore Ink 16 Deals, Advancing Cross-border Finance, Innovation and Cultural Ties

Viral Beijing Photo Brings Couple Back to Tie the Knot

