China’s progress in artificial intelligence is being hampered by U.S. trade restrictions on advanced chips, potentially widening the technological gap between the two nations, experts suggest.
During a discussion with Jefferies analysts, experts noted that despite high demand for Nvidia’s H20 chip, which bypasses the need for an export control license, trade restrictions are impeding China’s development of large language models (LLMs) that are integral to AI technologies such as ChatGPT. The analysts concluded that the gap between China and the U.S. is likely to persist or even grow.
The H20 chip offers less computing power compared to Nvidia’s chips available to U.S. companies. Consequently, Chinese firms are increasingly relying on Huawei’s Ascend chip as a long-term alternative. This move is anticipated to aid Huawei in enhancing its capabilities, although experts expressed concerns about China’s ability to sustain demand with the 7-nanometer process technology in the long term.
Reports indicate Huawei is struggling to ramp up production of its Ascend 910B chip—China’s top alternative to Nvidia’s restricted chips—due to the failure of components in modified chip fabrication equipment.
U.S. officials argue that Huawei’s Kirin 9000s chip, which uses 7-nanometer technology and powers the Mate 60 Pro smartphone, lags behind U.S.-developed chips in terms of sophistication. Secretary of Commerce Gina Raimondo asserted that the effectiveness of export controls is evident, as the Kirin 9000s is considerably less advanced than U.S. counterparts.
While the U.S. has recently reduced federal research and development funding across various sectors, China has increased its research and development investment by 10%, Arati Prabhakar, director of the White House Office of Science and Technology Policy, mentioned in an interview. She emphasized that amid the AI surge, the U.S. should be intensifying its efforts to stay competitive.
Experts agree that the demand for computing power will continue to rise as more advanced models are developed. They also predict consolidation in China’s AI industry due to the limited availability of capital and computational resources. This shortfall could weaken China’s competitiveness, with some experts suggesting that Chinese companies should optimize models to operate effectively with less computing power. However, another expert noted that initial AI model development stages require significant computing power for experimentation, and optimization is more applicable once models reach maturity.