Advanced Micro Devices (AMD) CEO Lisa Su delivered a bullish forecast for the future of artificial intelligence during a high-profile appearance in Shanghai, predicting that global AI adoption is on the verge of an unprecedented expansion. Speaking at AMD’s AI Developer Day, Su projected that approximately five billion people worldwide will interact with AI on a daily basis by 2030, fundamentally reshaping the global demand for computing infrastructure.
According to TechNode, Su emphasized that the industry has reached a critical inflection point. She firmly dismissed notions that the current AI boom is a transient bubble, stating that artificial intelligence is a foundational technology that will drive sustained, long-term demand for advanced semiconductors and data center infrastructure.
AI is not hype, Su told the audience of developers and industry partners. The sheer scale of her prediction, five billion daily users, implies that AI will soon become as ubiquitous as the internet or mobile connectivity, embedded seamlessly into enterprise workflows, consumer applications, and edge devices.
The 100x Computing Surge
To support this massive user base, Su forecast a dramatic surge in global computing requirements. She estimated that overall computing demand could increase 100-fold over the coming years, reaching the staggering scale of 10 yottaFLOPS (a yottaFLOPS is one septillion, or 10^24, floating-point operations per second).
This exponential growth in demand presents both a massive opportunity and a profound challenge for semiconductor designers like AMD and Nvidia, as well as foundry operators like TSMC. Meeting this demand will require not only scaling up the production of current-generation chips but also achieving significant breakthroughs in chip architecture, packaging, and energy efficiency.
The pressure to deliver this computing power is already reshaping the industry. As global chip equipment sales hit a record $135 billion in 2025, the race to build out fabrication capacity is accelerating.
AMD’s Commitment to Greater China
Su’s presence in Shanghai and her emphasis on AMD’s local footprint highlight the complex geopolitical tightrope that U.S. semiconductor companies must walk. Despite stringent U.S. export controls designed to limit China’s access to the most advanced AI accelerators, the Chinese market remains crucial for global chipmakers.
During her address, Su highlighted AMD must continue to develop in China’s “vibrant AI ecosystem.” She noted that the company employs more than 4,000 engineers across its major research and development centers in Greater China. Furthermore, AMD has established dedicated AI hubs in key technological centers, including Beijing, Shanghai, Shenzhen, and Taipei.
This extensive R&D presence underscores that China is not merely a consumer market for AMD, but a vital node in its global engineering and development network. However, operating in this environment is increasingly fraught. Earlier this year, China banned foreign AI chips from state-funded data centers, ordering the removal of Nvidia and AMD hardware in favor of domestic alternatives.
This push for self-reliance has accelerated the development of homegrown silicon. Chinese domestic chipmakers have already seized 41% of the local AI market as Nvidia’s grip loosens. Companies like Huawei and Alibaba are rapidly deploying their own accelerators, such as the newly announced Zhenwu M890 chip, to meet the surging domestic demand for computing power.
Navigating the Geopolitical Divide
For AMD, the challenge is to maintain its market share and leverage its Chinese engineering talent while strictly adhering to Washington’s evolving export regulations. The U.S. government has continually tightened the screws, with lawmakers recently introducing the MATCH Act to block AI chipmaking equipment sales to China.
Despite these headwinds, Su’s Shanghai address was decidedly focused on the technological horizon rather than the geopolitical present. By projecting a future where five billion people rely on AI daily, she painted a picture of a market so vast that it will require the combined efforts of the global semiconductor ecosystem to satisfy.
Whether AMD can fully participate in the Chinese segment of that 10-yottaFLOPS future remains an open question, dependent as much on policy decisions in Washington and Beijing as on the engineering prowess of its Shanghai-based teams. The company’s ability to sell its most advanced AI accelerators into China has already been curtailed by successive rounds of U.S. export controls, and further restrictions remain a constant threat.
Nevertheless, Su’s Shanghai appearance itself carries a message: AMD intends to remain a visible and active participant in China’s AI ecosystem for as long as regulations permit. By investing in local R&D talent, hosting developer events, and cultivating relationships with Chinese software engineers, AMD is building a foundation of goodwill and technical integration that could prove durable even as the hardware trade war intensifies. Su’s message was clear: the AI revolution is only just beginning, and the demand for computing power is about to enter uncharted territory.