Huawei is seeking to grab a larger share of China’s artificial intelligence chip market dominated by Nvidia, helping local companies adopt its rival’s silicon for so-called “lockdown” tasks.
Leading AI companies in China rely on graphics processing units (GPUs) made by Nvidia to “train” large language models, and the $3.4 billion US chipmaker’s products are seen as key to developing the technology.
Instead of challenging Nvidia in training, Huawei is positioning its latest Ascend AI processors as the hardware of choice for Chinese groups running “inference,” the computation LLMs undertake to generate an answer to a query.
The Chinese tech giant is betting that inference will be a bigger source of future demand if the pace of model training slows and AI applications like chatbots become more widespread.
“Training is important, but it only happens a few times,” said Georgios Zacharopoulos, a senior artificial intelligence researcher working on inference acceleration at Huawei’s Zurich lab. “Huawei is mainly focused on inference, which will ultimately serve more customers.”
It’s focused on the technically less challenging but potentially lucrative path of retrofitting AI models trained on Nvidia products to run on Ascend chips, according to company employees and Ascend customers. Since the Nvidia GPU and Ascend run on different softwareHuawei is helping companies use another software tool to make the two systems compatible.
Huawei’s push comes with top-down government support. Chinese officials have urged local tech giants to buy more of Huawei’s AI chips and move away from Nvidia.
One person familiar with Nvidia’s China operations said Huawei is internally seen as the country’s most serious competitor, adding that its chip design capabilities are “advanced.”
Washington has sought to curb the development of artificial intelligence in Beijing with export controls aimed at hindering the development of sensitive technologies in China.
Unlike its American rivals such as OpenAI and Google, the companies cannot access the most advanced GPUs in China. But even though Chinese groups can only source Nvidia’s smaller H20 chips that are fit for export controls, the less powerful GPUs are still in high demand because they are seen as better than local alternatives.
Analysts and Huawei researchers said Ascend isn’t yet ready to replace Nvidia for model training because of technical issues, such as a glitch in the way chips communicate with each other within a wider “cluster” of AI chips when training increasingly large models.
“Although the Ascend chips perform well on a chip-by-chip basis, there is a bottleneck with the connectivity between the chips,” said Lin Qingyuan, a Chinese semiconductor analyst at Bernstein. “When you train a large model, you have to break it down into smaller tasks. If one chip fails, the software needs to figure out a way for other chips to take over without delay.”
Another challenge for Huawei is convincing developers to switch from Nvidia’s Cuda software, known as the company’s “secret sauce” because it’s easy for developers to use and capable of significantly speeding up data processing.
But Huawei’s soon-to-be-released updated version of its AI chip, the Ascend 910C, is also expected to address these issues. “We expect this new generation of hardware to come with improved software that makes it more accessible to developers,” said a Huawei employee, who did not want to be named.
Huawei and Nvidia face stiff competition. Chinese internet group Baidu and chip designer Cambricon have made strides in developing AI chips. Meanwhile, in the US, Amazon and Microsoft are also betting that they can grab a larger market share in inference chips as AI applications become more widespread.
Estimates by SemiAnalysis, a chip consultancy, suggest that Nvidia made $12 billion in sales in China last year by shipping 1 million of its H20 chips to the country, selling twice as many AI chips as Huawei with its Ascend 910B.
“Nvidia’s China-specific H20 GPUs make up the majority of AI chips sold in China. But the lead is shrinking fast as Huawei ramps up production capacity,” said Dylan Patel, principal analyst at SemiAnalysis.
Industry insiders warned that Huawei’s push for AI chips was also limited by insufficient supply, with two potential buyers telling the Financial Times that they were unable to secure the chips.
Huawei did not respond to a request for comment. Nvidia declined to comment.
Analysts say Huawei’s production is likely to face challenges due to US export controls that have made Chinese factories dependent on outdated chip-making equipment.
The focus on inference also points to a different dynamic of Chinese AI development than America’s. Washington’s export controls mean Chinese AI players are not in the same race as Silicon Valley rivals Meta, Elon Musk’s x.AI and OpenAI to build large mega-clusters of Nvidia’s most advanced GPUs.
“Chinese companies are playing a different game. They pay much more attention to inference than the US because it is possible to achieve large efficiency gains even with less powerful chips, which also means they can achieve commercialization faster,” Bernstein analyst Lin said.
Chinese companies are betting they can stay competitive on AI by lowering inference costs, which in turn makes it cheaper to launch AI applications, he said.
Last month, Hangzhou- and Beijing-based start-up DeepSeek released its V3 model, which attracted attention for its low training and inference costs compared to comparable models in the US.
The company proposed a new way for an AI model to selectively focus on certain parts of the input data as a way of reducing the cost of running the model. It also used the “Mixture of Experts” technique popular among other Chinese AIs newly established companieswhich also helps speed up inference as only part of the model is used to generate the answer.
DeepSeek said Huawei successfully customized the V3 on the Ascend, providing detailed instructions for developers on how to use the chip. FT is earlier reported that Huawei has sent engineers to help users migrate from Nvidia to Ascend.
Additional reporting by Zijing Wu in Hong Kong