By Zoey
The contrasting visits to China by Lisa Su and Jensen Huang over the past week highlighted two sharply different approaches to navigating one of the world’s most politically sensitive and strategically important technology markets.
While Huang attracted public attention in Beijing with highly visible appearances, photo opportunities and enthusiastic crowds, Su adopted a far more restrained approach during meetings with customers, developers and Chinese officials, underscoring how the two U.S. chipmakers are responding differently to mounting geopolitical pressure and evolving market realities in China.
The diverging strategies come as the artificial intelligence chip market in China undergoes dramatic change due to escalating U.S. export controls, Beijing’s push for technological self-sufficiency and intensifying competition from domestic Chinese firms such as Huawei Technologies.
Just one year ago, Huang publicly acknowledged that Nvidia Corporation’s share of China’s AI chip market had dropped from approximately 95% to around 50% because of U.S. restrictions on advanced semiconductor exports.
This year, however, Huang said Nvidia’s effective share of the advanced AI accelerator market in China has essentially fallen to zero as stricter export rules blocked sales of the company’s most powerful chips and Chinese customers increasingly turned toward domestic alternatives.
The Nvidia chief has repeatedly warned that Washington’s restrictions risk permanently handing China’s AI market to local competitors. Huang previously estimated that China’s AI chip market alone could be worth roughly $50 billion this year, making it one of the most important battlegrounds in the global semiconductor industry.
By comparison, Advanced Micro Devices, better known as AMD, maintains a far smaller but more diversified presence in China.
Research firm IDC estimates AMD currently holds roughly 4% of China’s AI chip market. However, analysts say AMD benefits from broader access to the Chinese market because its business extends beyond AI accelerators into central processing units, gaming graphics processors, programmable chips and enterprise computing systems.
Lian Jye Su, chief analyst at research firm Omdia, said AMD’s wider product portfolio gives the company more flexibility in China as artificial intelligence applications spread into different sectors of the economy.
AMD can still supply Chinese customers with CPUs, consumer GPUs, field programmable gate arrays and various AI-related chipsets, allowing the company to participate in a broader range of computing infrastructure projects even under export restrictions.
Analysts say that advantage may become increasingly important as AI usage expands beyond training massive language models into enterprise software, industrial systems and commercial applications requiring more diverse hardware architectures.
The different approaches by AMD and Nvidia have become even more noticeable as both companies prepare to gather again in Taiwan next week for the annual Computex technology exhibition, one of Asia’s most influential events for the semiconductor and computing industries.
Both firms recently announced major investment plans in Taiwan, home to Taiwan Semiconductor Manufacturing Company, the world’s largest contract chipmaker and a critical supplier to both AMD and Nvidia.
The rivalry between the two executives carries an unusual personal dimension as well. Both Huang and Su were born in Taiwan and have publicly acknowledged that they are distant relatives, a detail that has drawn attention within the technology industry as their companies compete fiercely in the global AI race.
During her trip to China last week, Lisa Su focused primarily on customer engagement and developer outreach.
She spoke at an AMD developer conference in Shanghai and held meetings with Chinese clients and business partners. Her visit also included a publicly announced meeting with Chinese Vice Premier He Lifeng.
According to official statements, He told Su that China welcomes companies such as AMD to continue expanding cooperation and participating in the country’s economic development.
The high-level meeting stood in contrast to the comparatively muted official reception received by Jensen Huang during his own visit to Beijing earlier this month.
Huang was in China around the time of a major summit involving Chinese President Xi Jinping and U.S. President Donald Trump, yet Chinese authorities did not publicly announce an equivalent senior-level meeting for the Nvidia chief executive.
Industry observers said the difference may reflect Beijing’s careful balancing act as tensions between Washington and Beijing increasingly reshape global technology supply chains.
An adviser to multinational chief executives operating in China said Su’s quieter approach appeared better suited to the current geopolitical climate, where public scrutiny surrounding U.S.-China business relations has intensified.
The adviser noted that political sensitivities and growing nationalism in both countries have increased reputational risks for executives leading major global technology firms.
Despite the geopolitical challenges, China remains critically important to AMD’s business.
Su said last week that China accounts for approximately 20% of AMD’s overall revenue and remains a major market across personal computers, gaming hardware and certain data center products.
“Frankly, you look at the size of the market and the size of our portfolio, and we’ll continue to partner very closely with our Chinese customers,” she said.
At the same time, AMD is attempting to position itself as a viable alternative to Nvidia for Chinese AI developers seeking access to advanced computing technology.
During AMD’s Shanghai developer conference on May 19, Su strongly promoted ROCm, the company’s open-source software platform designed to compete with Nvidia’s dominant CUDA ecosystem.
The effort reflects growing demand among Chinese technology firms for alternatives as U.S. export restrictions make Nvidia’s most advanced hardware increasingly difficult to obtain.
Still, analysts caution that AMD faces significant challenges in expanding its influence in China’s AI sector.
Nvidia’s software ecosystem remains far more mature and deeply integrated into global AI development workflows, giving the company a substantial competitive advantage despite export restrictions.
A person familiar with AMD’s China operations said the company sold a significant number of AI chips to Alibaba Group last year. However, the source said Alibaba had to devote extensive engineering resources to adapting and debugging AMD systems because the software environment remained less developed than Nvidia’s.
Neither AMD nor Alibaba publicly commented on the matter.
As U.S.-China tensions continue reshaping the global semiconductor industry, analysts say the competition between AMD, Nvidia and rapidly advancing Chinese firms is likely to intensify further.
For now, however, the contrasting China strategies of Lisa Su and Jensen Huang reveal how even the world’s largest technology companies are being forced to navigate an increasingly complex intersection of business, geopolitics and artificial intelligence competition.