China’s development in artificial intelligence is being hampered by U.S. trade restrictions on advanced chips, a situation that could widen the technological gap between the two countries, according to experts.
Despite high demand for Nvidia’s H20 chip—which was designed to circumvent export control licenses—trade restrictions are limiting China’s ability to enhance its large language models (LLMs), critical for AI technologies like ChatGPT. This was highlighted during a discussion with Jefferies analysts, leading them to conclude that the gap between China and the U.S. in AI capabilities may remain or even expand.
The H20 chip has less computing power compared to Nvidia’s other chips available in the U.S., prompting Chinese companies to look toward Huawei’s Ascend chip as a long-term alternative. Experts noted that this shift could also aid Huawei in improving its chip technology. However, there is concern about China’s capacity to produce enough of these advanced chips using the 7-nanometer process to meet future demands.
Huawei is reportedly facing challenges in ramping up production of its Ascend 910B chip, one of China’s best alternatives to Nvidia’s restricted chips, due to issues with components in repurposed chip fabrication machines.
U.S. officials have stated that another chip, the Kirin 9000s—which relies on advanced 7-nanometer process technology and powers Huawei’s Mate 60 Pro smartphone—is still behind the technology used in chips developed in the U.S.
Secretary of Commerce Gina Raimondo commented that export controls appear to be effective, as the Kirin 9000s chip is far less advanced than those produced in the United States. “We have the most sophisticated semiconductors in the world. China doesn’t,” she said.
Additionally, Arati Prabhakar, director of the White House Office of Science and Technology Policy, pointed out that while the U.S. has recently reduced federal research and development funding across various industries, China has increased its research and development spending by 10%. In an interview with The Verge, Prabhakar emphasized that as AI technology grows, the U.S. should be increasing its efforts, not cutting back, to maintain its edge.
Looking ahead, experts predict a continued rise in demand for computing power as more advanced AI models are developed. They also foresee a potential consolidation in China’s AI industry due to the limited availability of both capital and computing power. This shortage could diminish China’s competitiveness, they warned. While some experts suggested that Chinese companies should focus on optimizing models to perform competitively with less computing power, another expert noted that the initial stages of AI model development require significant computing power and that optimization would only be effective once the models are more mature.