China’s development in artificial intelligence is being hindered by U.S. trade restrictions on advanced semiconductor chips, which could widen the technological gap between the two nations, according to experts.
Despite high demand for Nvidia’s H20 chip, designed to bypass export control licenses, trade restrictions are limiting China’s ability to enhance its large language models (LLMs) that power AI technologies like ChatGPT. A note from analysts at Jeffries indicated that this could result in China’s technological gap with the U.S. either remaining the same or expanding.
The H20 chip, offering lower computing power than Nvidia’s chips available to U.S. tech companies, has pushed Chinese companies to consider Huawei’s Ascend chip as a long-term alternative. Experts noted that this shift could also help Huawei improve its chips, though there are concerns regarding China’s capacity for advanced 7-nanometer process to meet long-term demand.
Reports suggest Huawei is struggling to increase production of its Ascend 910B chip, considered China’s best alternative to the restricted Nvidia chips, due to issues with components in repurposed chip fabrication machines.
U.S. officials have pointed out that another Huawei chip, the Kirin 9000s, which employs 7-nanometer processing technology and powers Huawei’s Mate 60 Pro smartphone, is not as advanced as those being developed in the U.S.
“These export controls are effective because that chip is significantly behind what we have in the United States,” said Secretary of Commerce Gina Raimondo. “We possess the most advanced semiconductors in the world. China does not.”
Despite these restrictions, China has increased its research and development funding by 10%, compared to cutbacks in U.S. federal research and development across various industries. Arati Prabhakar, director of the White House Office of Science and Technology Policy, stated that the U.S. should be “doubling down” on its efforts during the AI boom.
Experts predict that the demand for computing power will continue to rise as companies create larger and more sophisticated models. They also mentioned the likelihood of consolidation within China’s AI industry due to current shortages in capital and computing power.
This shortage could impact China’s global competitiveness in AI. While some suggest that Chinese firms should focus on optimizing existing models to perform with less computing power, it was also noted that the initial stages of AI model development require substantial computing power for experimentation, with optimization becoming more beneficial as models mature.