As one of the leaders in AI technology, the United States attaches great importance to its competitive advantage in the global market. However, when the Chinese market rises and becomes the world's largest AI chip market, the United States can't help but feel anxious. In particular, NVIDIA, as a company that occupies an important position in the field of AI chips, has become even more worried about the development momentum of the Chinese market. Therefore, the U.S. Secretary of Commerce said that NVIDIA will not be allowed to provide China's advanced AI chips, hoping to hinder the development of China's AI technology and maintain the leading position of the United States in science and technology by restricting the sale of chips.
However, the U.S. blocking measures have not discouraged Chinese AI chip companies, on the contrary, they have accelerated the pace of research and development. A number of Chinese AI chip companies have announced that their chip performance is close to that of NVIDIA's A100 chip, a breakthrough that highlights the rapid development of AI chips in China. At the same time, Chinese chip companies are also able to deeply optimize AI software because of their mastery of the underlying technology of chips, making it more outstanding in terms of performance and energy saving. In contrast, even if NVIDIA launches a customized A800 chip, its performance lags behind that of China's AI chips. Now the United States does not even allow NVIDIA to emasculate the A800 chip for China, which will only further prompt domestic companies to speed up the substitution process of domestic AI chips.
The growing demand for AI chips in the Chinese market has not only attracted the attention of foreign chip companies such as NVIDIA, but also spawned the rise of domestic companies. In the second half of this year, a Chinese Internet giant purchased a large number of domestic AI chips, and the number of orders far exceeded that of NVIDIA. This shows that domestic companies have begun to replace imported AI chips on a large scale and rely more on domestic AI chips. Moreover, because the performance of AI chips that NVIDIA can sell is lower than expected, it can only reach 20% of the performance of A100, which forces domestic companies to accelerate the substitution process of domestic AI chips.
However, at present, the main problem in the substitution of domestic AI chips is ecological construction. NVIDIA's CUDA ecosystem, which it has built over the years, has become its biggest competitive advantage. In order to break the monopoly of CUDA, domestic AI chip companies provide chips to domestic universities for free, so that professors and students can familiarize themselves with domestic AI chips, and gradually establish a domestic AI chip ecosystem. Even the American chip company Intel has called for breaking NVIDIA's ecological monopoly, and Chinese chips are expected to become the object of Intel's cooperation. If Chinese and American chip companies can cooperate, then NVIDIA's CUDA ecological monopoly will usher in a rapid end.
In general, the United States is trying to keep American chips from continuing to earn huge revenues from the Chinese market, while also preventing the development of Chinese high-tech and fooling Chinese companies with backward AI chips. However, Chinese AI chips have made a breakthrough and accelerated the substitution of imported chips. When China's AI chip ecosystem is established, NVIDIA's last reliance will also be lost, and at that time, NVIDIA will no longer have a chance in the Chinese market. That's what worries Nvidia the most. As a result, NVIDIA founder Jensen Huang has been calling for allowing him to deliver AI chips with stronger performance to China, hoping not to lose opportunities in the Chinese market.
This article introduces the situation of the United States blocking Nvidia's high-performance AI chips in China, and analyzes the fear of the United States and the substitution momentum of China's AI chips. Although the United States has tried to hinder China's AI technology development by restricting chip sales, Chinese AI chip companies have made breakthroughs, and the performance is close to or even better than that of NVIDIA's chips. Moreover, Chinese chip companies have mastered the underlying technology of chips and are able to optimize AI software to make it more competitive. At the same time, the demand for AI chips in the Chinese market is growing, and domestic companies are also accelerating the substitution process of imported chips. While NVIDIA has an advantage by virtue of its established CUDA ecosystem, Chinese chip companies are committed to breaking the monopoly of CUDA by partnering with universities. Moreover, even the American chip company Intel has expressed its desire to break the monopoly of the NVIDIA ecosystem, and Chinese chips are expected to become its partners. In general, it is difficult for the United States to achieve its dual goal of continuing to earn income from the Chinese market while preventing the development of China's high technology. Once China's AI chip ecosystem is established, NVIDIA will lose opportunities in the Chinese market, which is also the thing that NVIDIA is most worried about. Therefore, the rise of China's AI chips is unstoppable, and its position in the domestic market will continue to improve, which will have a significant impact on the international competition pattern.