The controls pertain to A100 and H100 chips, intensifying US measures to restrict China’s access to these products
The United States has extended export restrictions on Nvidia artificial intelligence chips from just China to include select Middle Eastern countries. In a recent regulatory filing, Nvidia, valued at $1.2 trillion, disclosed that these restrictions impact their A100 and H100 chips, integral for enhancing machine learning operations in prominent AI applications, like ChatGPT.
The company stated that these controls would not result in an “immediate significant effect” on its outcomes. However, it did not specify the Middle Eastern countries affected by these limitations.
It was reported by an anonymous source to Reuters that AMD, a competitor in the same sector, had also received a notification letter outlining similar restrictions.
Nvidia issued a statement, explaining, “In the second quarter of fiscal year 2024, the US government apprised us of an added licensing mandate applicable to specific customers and regions, including select Middle Eastern countries, for a subset of A100 and H100 products.”
Notably, the sales of A100 and H100 chips are already prohibited in China and Russia. The decision to expand trade restrictions signifies an escalation in the Biden administration’s endeavors to limit Beijing’s capacity to benefit from the AI advancement.
In August of the previous year, Nvidia disclosed that U.S. authorities had instructed them to halt chip exports to China. Officials conveyed that the new export regulation aimed to mitigate the potential risk that the specified products could be utilized in or redirected toward a “military end use” or “military end user” within China.
In October, the U.S. introduced additional export controls with the objective of severing China’s access to specific semiconductor chips, regardless of where they were manufactured but involving U.S. equipment.
High-ranking government officials clarified that many of these regulations were designed to obstruct foreign companies from selling advanced chips to China or providing Chinese entities with the tools required for manufacturing advanced chips on their own.
Deprived of chips from firms like Nvidia and AMD, Chinese entities will encounter challenges in cost-effectively conducting advanced computing tasks, including image and speech recognition, among numerous others.
Nvidia’s CEO, Jensen Huang, cautioned back in May of this year that the United States faced the prospect of causing “significant harm” to its technology sector if it persisted with trade restrictions on China.
During an interview with the Financial Times, Huang stated, “If [China] cannot purchase from the United States, they will simply develop it domestically. Therefore, the U.S. needs to exercise caution. China is an exceptionally crucial market for the technology sector.”
Just earlier this month, Nvidia announced quarterly revenue of $13.5 billion, surpassing predictions by $2 billion.
