In a recent development, Nvidia, the renowned American multinational technology company, has voiced its concerns over the potential repercussions of the US government's decision to impose further semiconductor export restrictions to China. The company warns that these curbs could lead to a "permanent loss" for the US semiconductor industry in one of the world's most significant and fiercely competitive markets.
Nvidia's apprehensions stem from the regulations introduced last year, which prohibited the company from selling its high-end A100 and H100 graphics processing units to China. Recent reports suggest that Washington is contemplating the imposition of new export restrictions on semiconductors used in artificial intelligence (AI), a domain where Nvidia is a key player.
During the company's earnings call on Wednesday, August 23, Nvidia's CFO, Colette Kress, stated, "We believe the current regulation is achieving the intended results." She further added that while the immediate financial impact of these restrictions on Nvidia's data center GPUs might not be significant, the long-term implications could be detrimental. The restrictions could erode the US semiconductor industry's competitive edge in the Chinese market.
This statement from Nvidia echoes its warning from June, where the company highlighted the potential adverse effects on the US semiconductor industry due to further curbs on chip exports.
Financial data reveals that China contributes to 20% to 25% of Nvidia's revenue in its data center business, which is currently its most significant unit. This unit has witnessed a staggering 171% year-on-year sales growth, culminating in a record revenue of $10.32 billion in the June quarter.
The US government's rationale behind these export restrictions is to prevent China from accessing technology that could potentially be repurposed for military objectives. This is in light of China's ambitions related to Taiwan and the South China Sea, and its broader geopolitical aspirations in the Asia-Pacific region and potentially globally.
Despite the geopolitical tensions between the US and China, Nvidia's chips, especially those used in data centers for training extensive AI models, are considered superior to their competitors. Chinese firms, which rely heavily on such chips for training vast data sets, consume approximately 40% of the global chip production. However, their self-sufficiency in this domain stands at a mere 12%.
The ongoing tussle between the US and China over chip exports underscores the broader challenges and complexities of the global tech industry, where geopolitical considerations often intersect with business interests. As the situation unfolds, companies like Nvidia find themselves at the crossroads of policy decisions and market dynamics.
