The Trump administration has recently imposed restrictions on Nvidia Corporation, preventing the company from selling its H20 chip in China. This move marks a significant escalation in the ongoing tech battle between Washington and Beijing, resulting in substantial financial losses for Nvidia and hindering a product line that was specifically designed to comply with previous US regulations.
On Monday, the US government notified Nvidia that exporting the H20 chip to China would require a license “for the indefinite future.” This decision was made due to concerns that the chip could potentially be used in Chinese supercomputers. As a result, Nvidia anticipates reporting approximately $5.5 billion in writedowns in the current quarter related to inventory and commitments for the H20 chip.
Following this announcement, Nvidia’s shares fell by 6.9% in New York, triggering a broader selloff in semiconductor companies across the US and Japan. South Korean chipmakers Samsung Electronics Company and SK Hynix also experienced a decline of around 3%, while Advanced Micro Devices (AMD) stated that it expects to incur a charge related to the new restrictions on its products.
Additionally, Dutch company ASML Holding reported disappointing financial results for the first quarter, further exacerbating the situation in the semiconductor industry. The uncertainty surrounding tariffs has added to the challenges faced by chip manufacturers, impacting their performance in the market.
The Commerce Department has issued new export licensing requirements for Nvidia’s H20 chip, AMD’s MI308, and similar products. These measures are part of the government’s efforts to safeguard national and economic security, aligning with President Trump’s directive to limit China’s semiconductor and artificial intelligence ambitions.
The restrictions imposed on Nvidia and other AI chipmakers in China are a continuation of previous actions taken by the US government to address concerns about technology transfer and national security risks. Nvidia developed the H20 chip in response to earlier restrictions, aiming to provide a product that complies with regulatory requirements while still offering AI capabilities.
Under the Biden administration, regulations were considered that would have impacted Nvidia’s H20 and B20 chips, potentially increasing the cost of developing AI models in China. However, these measures were not pursued, as the administration focused on implementing broader AI chip restrictions that would affect global markets.
Nvidia has expressed opposition to the proposed AI diffusion rule, arguing that it could disadvantage American companies and strengthen China’s technological independence. Trump’s team is looking to enhance and streamline these measures, though the specifics of these efforts remain unclear. Companies are expected to comply with the global restrictions within a month.
The impact of the H20 restrictions on Nvidia’s revenue could be substantial, potentially resulting in missed opportunities for the company. Analysts predict that if the restrictions persist, Nvidia’s data-center exposure to China may return to levels seen in early 2024, before the production of the H20 chip ramped up.
Despite these challenges, Nvidia recently announced plans to invest in AI infrastructure in the US, signaling its commitment to the market. The company’s decision to build AI infrastructure worth up to $500 billion over the next four years reflects its ongoing efforts to expand its presence in the industry.
In conclusion, the restrictions on Nvidia’s H20 chip highlight the complexities of the ongoing tech battle between the US and China. As the semiconductor industry continues to evolve, companies like Nvidia will need to navigate regulatory challenges while pursuing innovation and growth opportunities in the global market.
