IN Declaration to the Securities and Exchange Commission Last Friday, Nvidia announced that US government officials had ordered restrictions on sales of its top AI chips in China and Russia. The new restrictions (in the form of licensing requirements subject to US government approval) include the powerful A100 Tensor Core GPUthe upcoming H100and any chips of equivalent power or systems incorporating them. The goal is to “address the risk that covered products could be used or diverted to ‘military end use’ or ‘military end user’ in China and Russia,” according to Nvidia, which notes that the company already does not sell products to customers in Russia.
Reuters reports that the Commerce Department intends the new policy to “keep advanced technology out of the wrong hands.” China is unhappy with the restrictions, calling the move part of a “technology blockade”. The US also restricted AMD sales Accelerator MI250 AI chip for China. Whether this effort will have any effect on China’s AI capability in the long term remains to be seen, as Chinese firms have begun development of own GPUs to use graphics and AI.
Currently data centers packed with A100 GPUs powerful high-end AI training operations on image recognition systems and large language patterns; they are used for state-of-the-art training image synthesis models which can be used to generate propaganda campaigns, deep falsifications and disinformation. GPUs provide an ideal platform for developing neural networks because massive parallelism in their architecture, which dramatically speeds up the computation time compared to CPU-based methods.
Nvidia manufactures the A100 in Taiwan, a current geopolitical flashpoint between the US and China. The firm said in the SEC filing that the restrictions could affect the completion of its H100 project and “may require the company to transfer certain operations” outside the country. Wall Street reacts to news of AI chip limitations negative, which sent Nvidia and AMD stock prices down. Given US-China tensions over Taiwan and the rapid development and deployment of AI technologies in research and business, this is likely just the beginning of a much larger ongoing story.