The future of \u00a0Nvidia\u00a0 in \u00a0China\u00a0 is becoming increasingly uncertain. At the start of October 2024, the Chinese administration issued a \u00a0recommendation\u00a0 to artificial intelligence (AI) companies, urging them to prioritize the use of \u00a0chips produced in China\u00a0. Fast forward ten months, and this recommendation has evolved into a \u00a0requirement\u00a0. The Chinese government is now mandating that state-owned data centers utilize at least \u00a050%\u00a0 of Chinese integrated circuits in their servers.<\/p>\n
<\/p>\n
In contrast, Nvidia has finally acquired the export license necessary to sell its \u00a0H20 AI GPU\u00a0. However, the Chinese government has effectively vetoed this chip. The \u00a0Cyberspace Administration of China\u00a0, which regulates internet activity within the country, is currently investigating the H20 GPU due to suspicions that it could contain a \u00a0backdoor\u00a0 that might be exploited by foreign entities. This investigation severely limits the likelihood of China adopting this chip for its needs.<\/p>\n
<\/p>\n
The direct fallout from this unfavorable scenario for Nvidia is evident: in the last quarter, the company did not sell a single H20 GPU in China, as noted by \u00a0Shaun Rein\u00a0, an expert in the Chinese economy and founder of the \u00a0China Market Research Group (CMR)\u00a0 based in Shanghai. While this statement holds some truth, it does have a catch. For a considerable part of last quarter, Nvidia lacked the export license required to deliver this chip to Chinese clients, though it now possesses it. With this license in hand, it could have shipped \u00a0thousands\u00a0 of these GPUs in recent weeks.<\/p>\n
<\/p>\n
Despite aggressive efforts from the U.S. government to stifle competition, \u00a0cutting-edge AI chips\u00a0 are still finding their way into China. These chips are primarily being imported through \u00a0secondary markets\u00a0 and \u00a0parallel imports\u00a0 from countries like \u00a0India, Malaysia,\u00a0 and \u00a0Singapore\u00a0, where U.S. regulations hold limited sway. Moreover, developers creating significant AI models that utilize \u00a0CUDA\u00a0 have discovered reliable avenues through the international second-hand market.<\/p>\n
<\/p>\n
\u00a0Cambricon Technologies\u00a0 is a rising star among companies specializing in AI chip design.<\/p>\n<\/p><\/div>\n<\/div>\n
China currently boasts three significant alternatives to Nvidia. Among them, \u00a0Cambricon Technologies\u00a0\u2014while not as well-known as \u00a0Huawei\u00a0 or \u00a0Moore Threads\u00a0\u2014is one of the companies with the \u00a0highest growth potential\u00a0 in the design of AI GPUs. Recently, it secured approval from the \u00a0Shanghai Stock Exchange\u00a0 to \u00a0raise $560 million\u00a0 for the development of four chips destined for AI model training and inference, and to create an alternative to \u00a0CUDA\u00a0.<\/p>\n
<\/p>\n
Another key player, \u00a0Moore Threads\u00a0, has developed several GPUs targeted for AI applications that compete with advanced offerings from Nvidia, AMD, and Huawei. Noteworthy products in its lineup include the \u00a0MTT S4000\u00a0 and \u00a0MTT S3000\u00a0 cards, which offer capabilities that rival established market leaders. Interestingly, the \u00a0MTT S80\u00a0 card, designed for gaming and content creation, claims to deliver a computation capacity of \u00a014.4 TFLOPS\u00a0 in single-precision floating-point operations.<\/p>\n
<\/p>\n
The third essential player in China\u2019s AI chip sector is \u00a0Huawei\u00a0. Its flagship product, the \u00a0Ascend 910D\u00a0 chip, is designed to surpass the performance of Nvidia\u2019s \u00a0H100 GPU\u00a0. Huawei has also recently introduced the \u00a0Ascend 920\u00a0 chip, strategically aimed at filling the void that the Nvidia H20 GPU will leave in China\u2019s market. This chip is slated for large-scale production in the latter half of \u00a02025\u00a0, employing \u00a06nm\u00a0 integration technology purportedly developed in collaboration with \u00a0SMIC\u00a0 (Semiconductor Manufacturing International Corporation).<\/p>\n
<\/p>\n
Image | Nvidia<\/a> | Zhang Kaiyv<\/a><\/p>\n More information | Shaun Rein<\/a><\/p>\n In Xataka | The U.S. gives \u00a0Huawei\u00a0 a tremendous opportunity to fill the gap left by Nvidia in China\u2019s AI chip market.<\/p>\n