In 2025, the Huawei booth at the Mobile World Congress in Barcelona.
Arjun Kharpal | CNBC
Despite being overwhelmed by years of US trade restrictions, Chinese telecom giant Huawei has quietly appeared as one of the country’s most intense competitors across the AI landscape.
The Deep Shenzhen-based company not only appears to represent Beijing’s answer to the American AI chip darling Nvidia, but it is also the early adoption of monetization of artificial intelligence models in industrial applications.
“Huawei has been forced to shift and expand its core business focus over the past decade. Due to various external pressures on the company.”
With this expansion, the company is involved in everything from smart cars and operating systems to advanced semiconductors, data centers, chips, and large-scale language models, and more.
“Other technology companies have not been able to be competent in many different sectors as they have higher complexity and higher barriers to entry,” Triolo said.
This year, Nvidia CEO Jensen Huang is increasingly speaking up by calling Huawei “one of the world’s most frightening tech companies.” He also warns that if Washington continues to restrict exports of US chip companies to Asian countries, Huawei, replacing China’s Nvidia, will replace China’s Nvidia.
Nvidia surpassed its market capitalization last week, exceeding its $4 trillion market capitalization to become the world’s most valuable company. The state-of-the-art processors and associated “CUDA” computing systems continue to be the industry standard for training generative AI models and applications.
However, the moat may be narrower. Huawei is proving that it’s not just everything, it works. It’s tall orders that challenge American AI stubborns like Nvidia, but the company’s history shows why it’s not counting.
Switching phone calls to National Champion
Currently employing over 208,000 people in over 170 markets, Huawei has come from a humble beginning. Founded in 1987 by ambitious entrepreneur Ren Zhengfei from a deep Shenzhen apartment, the company started as a small phone switch distributor.
As it grew into a global communications player, it gained traction by targeting undeveloped markets such as Africa, the Middle East, Russia and South America, ultimately expanding to places like Europe.
By 2019, Huawei was well positioned to leverage global 5G rollouts and become a market leader. Around this time it also blossomed into one of the world’s largest smartphone manufacturers, designing smartphone chips through Hisilicon, a subsidiary of Chip Design.
However, Huawei’s success also attracted an increase in scrutiny from governments outside China, particularly the US. This frequently denies Huawei’s technology for poses a national security threat. Chinese companies are refuting such risks.
Export restrictions ironically pushed Huawei into the Chinese government’s weapons, in a way that CEO Ren Zhengfei always resisted.
Paul Triolo
China Partner and Senior Vice President of DGA-Albright Stonebridge Group
Huawei’s business suffered a major set-off when it was placed on the US trade blacklist in 2019, preventing American companies from doing business with it.
Once the impact of sanctions began, Huawei’s consumer business was once the largest along with our revenues – half of the previous year to about $34 billion in 2021.
The company still manages AI chip breakthroughs, and despite the addition of US restrictions in 2020, Huawei officially launched its Ascend 910 AI processing chip as part of its strategy to build a “full stack, all-senario AI portfolio power” despite cutting the company from chip maker Taiwan Semiconductor Manufacturing Co. a year ago despite additional US restrictions, Huawei officially launched its Ascend 910 AI processing chip as part of its strategy to build a “full stack, all-senario AI portfolio power.”
However, Huawei’s US targeting also helped transform the company into someone like a Chinese martial artist, attracting attention in 2018 when Huawei’s CFO and daughter Meng Wanzhou was arrested in Canada for violating Iran’s sanctions.
As the US-China tech war continued to widespread and advanced chipping restrictions were imposed on China, Huawei became Beijing’s AI superpower and was one of the beneficiaries of its push to achieve technical self-sufficiency.
“Export restrictions ironically pushed the star into the Chinese government’s weapons, just as CEO Len Chang Fei always resisted,” Triolo said. In this way, the limitation has also become “steroids” in Huawei’s AI hardware and software stack.
comeback
After further declines in sales in the consumer segment, the unit began to turn to 2023 with the release of a smartphone that analysts said included advanced chips made in China.
The 5G chip shocked many people in the US. He didn’t expect Huawei to reach that level of progress without TSMC. Instead, Huawei reportedly worked with Chinese chipmaker Smic.
Semiconductor analysts say the scale at which Huawei and Smic can generate these chips is severely limited, but Huawei has proven that it is back to advanced chip games.
Also, around this time, reports began to surface for Huawei’s new AI processor chip, the Ascend 910b. The company is considering seizing the gap left by Nvidia’s most advanced chip export control. It is said that mass production of the next-generation 910C is already on the way.
“We’re committed to providing a range of services to our customers,” said Jeffrey Towson, managing partner at TechMoat Consulting.
In April, Huawei announced the AI CloudMatrix 384, a system that links 384 Ascend 910C chips in clusters within data centers. Analysts say CloudMatrix can outperform Nvidia’s system, the GB200 NVL72, in some metrics.
Huawei “is not just redefine how AI infrastructure works,” Forrester analysts said in a report last month on CloudMatrix.
Meanwhile, Huawei is also developing its own “Cann” software system that will serve as an alternative to Nvidia’s CUDA.
“Winning AI races doesn’t just make chips faster, it also includes delivering the tools developers need to build and deploy large models,” the Forrester report said, but the authors pointed out that Huawei’s products are still not fully integrated with other commonly used tools for developers to switch quickly from NVIDIA.
“The rising ecosystem strategy”
Huawei’s goal to surpass Nvidia is seen as a major development in China and US competition for AI, but it is important to note that the chip represents one building block of Huawei’s broader AI plan.
Huawei currently has its hands across the artificial intelligence value chain, from chips to computing, AI models and AI applications. These various AI business instruments also utilize other areas of the company’s vast technology empire.
In 2024, revenues rose 22.4% to 862 billion yuan. Huawei’s “ICT Infrastructure” business (including the development of industrial AI systems) is the company’s largest revenue driver at 369.9 billion yuan, while its cloud computing business generated 68.8 billion yuan, an increase of 24.5% from the previous year.
The company deploys the growing AI data centers, operated by the cloud computing unit Huawei Cloud, Ai Chips and AI CloudMatrix 384. Amazon Web services and Oracle.

These data centers provided training capabilities and computing power that Huawei’s suite of AI models use under the Pangu series.
Unlike other general-purpose AI models such as Openai’s GPT-4 and Google’s Gemini Ultra 1.0, Huawei’s Pangu model is designed to support more industry-specific applications across the healthcare, finance, government, industry and automotive sectors. Pangu has already been applied in more than 20 industries last year, the company said last month.
Deploying such AI applications often involves Huawei Tech staff working for several months on project sites such as Vice President of Marketing for Remote Coal Mines, Huawei’s Oil, Gas and Mining Business Units.
The study allowed the company to deploy more than 100 power-supply trucks that could autonomously transport soil and coal using the electric company’s 5G network, AI and cloud computing services.
And it is not limited to China. The technology can be “recreated on a large scale in Central Asia, Latin America, Africa and the Asia-Pacific region,” Chen said.
Huawei also said it has open-sole the Pangu model and will help expand overseas and further expand its “Ascend Ecosystem Strategy,” which refers to an AI product built around Ascend chips.
Speaking to CNBC’s “Squawk Box Asia” on Thursday, Patrick Moorhead of Moor Insights & Strategy said Huawei is hoping to promote the country, part of China’s Belt and Road Initiative.
Over the course of five to ten years, the company was able to start building serious market shares in these countries.