Intel tightens AI chip exports to China

[WORLD] AMERICAN chipmaker Intel has informed Chinese clients it will start needing a license to sell some of its advanced artificial intelligence processors, the Financial Times reported on Wednesday.

The news comes a day after Nvidia warned of a $5.5 billion hit as Washington restricted exports of its AI processor tailored for China. Dutch chip-making tools giant ASML also raised doubts about its outlook earlier in the day.

Intel's decision follows a series of export restrictions imposed by the U.S. government on advanced semiconductor technologies, aimed at curbing China's rapid advancements in artificial intelligence and high-performance computing. These restrictions have sparked concerns among industry experts about the potential disruption to global supply chains and the broader impact on technological innovation.

Intel, under new CEO Lip-Bu Tan, told clients last week that its chips would require a license for exporting to China if they have a total DRam bandwidth of 1,400 gigabytes (GB) per second or more, input-output (I/O) bandwidth of 1,100 GB per second or more, or a total of both of 1,700 GB per second or more, according to the report.

This latest development comes as Washington intensifies its efforts to limit China's access to semiconductors with military and AI applications. The U.S. Commerce Department has steadily expanded its list of controlled technologies, placing additional pressure on companies like Intel, Nvidia, and AMD to comply with evolving export rules. In response, China has ramped up investments in domestic chip development through state-backed initiatives like the “Made in China 2025” plan.

The move by Intel reflects the growing complexity of navigating geopolitical tensions in the tech industry. While Intel aims to comply with U.S. regulations, it also faces the challenge of maintaining its market position in China, which is a significant consumer of advanced semiconductor products. This delicate balance is further complicated by the evolving nature of U.S. trade policies under the Trump administration.

Intel's Gaudi series as well as Nvidia's H20 far exceed these requirements, the report said. The chipmaker did not immediately respond to a Reuters request for comment.

Several Chinese tech firms, including Tencent, ByteDance, and Alibaba, are said to be reassessing their AI infrastructure roadmaps amid fears of long-term hardware shortages. According to analysts at Bernstein, some companies have already begun stockpiling older-generation chips and seeking alternatives from domestic suppliers, although these often fall short in performance and efficiency.

ASML, another major player in the semiconductor industry, has also been affected by these trade tensions. Despite reporting a 92% increase in net profit for the first quarter of 2025, ASML expressed concerns about the impact of U.S. tariffs on its future business. The company's executives noted that while artificial intelligence remains a core driver of industry growth, the macroeconomic uncertainty introduced by trade policies could lead to fluctuations in demand.

The ripple effects of these restrictions are also being felt in global R&D investment. Several multinational firms are reportedly scaling back their collaborative research projects with Chinese institutions, fearing future sanctions or compliance violations. Experts warn that this decoupling could slow the pace of innovation in key areas like quantum computing, 5G, and neural network optimization.

Intel's shares closed more than 3% lower on Wednesday, under pressure like other chip stocks, on fresh evidence of how U.S. President Donald Trump's shifting trade policy was complicating the outlook for semiconductor and computing giants.

The AI chip trade is losing steam after a two-year rally as tariff threats and fears over Big Tech's spending weigh on sentiment.

Despite these headwinds, some investors remain cautiously optimistic about long-term demand for AI hardware. Venture capital in the AI infrastructure space continues to grow, particularly in regions outside of the U.S.-China axis, such as the Middle East and Southeast Asia. This suggests that while trade restrictions may reshape supply routes, the global appetite for AI computing power is unlikely to disappear.