In a recent development, Nvidia has received limited authorization to continue its development and export of two advanced chips to China. This permission was granted by the Department of Commerce shortly after Nvidia disclosed that it was subject to new regulations requiring a special license for exporting these chips to China.
Under the newly granted authorization, Nvidia can transfer the necessary technology to China to develop its H100 chip until March 1, 2023. Additionally, the company can carry out order fulfillment and logistics for its H100 and A100 chips through its Hong Kong facility until September 1, 2023.
The Biden administration has been increasingly vigilant in scrutinizing U.S. chip suppliers amid efforts to tighten control over the export of sensitive technology to China. Nvidia’s announcement on August 26 revealed that the rule change was aimed at mitigating the risk of the covered products being used for “military end use” or “military end user” purposes in China and Russia. It’s worth noting that the licensing requirement also applies to Russia, although Nvidia ceased sales to the country in March.
The restrictions primarily targeted Nvidia’s exports of A100 and H100 chips, which are integral to artificial intelligence (AI) and other emerging technologies. Nvidia is a significant producer of these chips, which it both develops and sells in China.
These new licensing requirements have the potential to impact Nvidia’s sales to data centers in China, a crucial customer segment in the country. In its initial filing, Nvidia estimated that without any allowance, it could jeopardize up to $400 million in third-quarter sales.
Nvidia is actively working with its customers in China to fulfill planned or future purchases using alternative products and is considering seeking licenses where replacements are insufficient, as stated in their official statement.
The United States is increasingly proactive in enforcing export controls on critical technology as it seeks to safeguard the global use of its most advanced technology. Earlier in August, the Department of Commerce announced stricter controls on two types of advanced semiconductors to protect them from potential misuse.
The long-term impact of the government’s rule change for Nvidia on U.S. chip suppliers remains uncertain, and much will depend on the licenses ultimately granted. Karson Elmgren, a research analyst with Georgetown University’s Center for Security and Emerging Technology, noted that there are trade-offs in clamping down on exports, particularly concerning Nvidia’s position as a leading developer of crucial hardware for AI and other technologies. While additional restrictions in the industry are considered unlikely at this time, the situation remains dynamic.