US Sanctions Propel Chinese AI Prodigy to $23 Billion Fortune
(Bloomberg) — In 2019, Chen Tianshi was far from being one of the richest people on the planet.
Chinese telecommunications giant Huawei Technologies Co., the three-year-old artificial intelligence chip startup’s biggest customer, had abruptly cut nearly all of its business to develop its own semiconductors. By then, Huawei was the source of more than 95% of the company’s revenue.
But then he caught a breakout from an unexpected source. The U.S. decision to cut off China’s access to cutting-edge chips and Beijing’s determination to promote domestic technology ultimately created a halo of state sponsorship and a vast protected market for the computer genius’s company, propelling him to become one of the world’s richest self-made billionaires.
Shares of chip designer Cambricon Technologies are up more than 765% in the past 24 months. His fortune, mostly from his 28% stake in the Beijing-based maker of artificial intelligence accelerators, has more than doubled since the beginning of the year to $22.5 billion, according to the Bloomberg Billionaires Index.
Chen’s meteoric rise underscores how China’s staunch support for its domestic AI industry has created a new class of state-linked tech elites just years after crushing private sector giants. As Washington’s export bans clog China’s access to advanced chips, firms like Chen’s Cambricon have emerged as national champions, protected by policy mandates and investor zeal; these were symbols of a new industrial order in which political favor, not market freedom, defined the winners.
Questions about how much support from government protectionism, rather than the competitiveness of its chips, contributed to Cambricon’s rise have divided observers over how long this process will take.
“Cambricon’s explosive revenue growth is mainly driven by a low starting point, and its current valuation could inflate without sustained policy support,” said Shen Meng, director at Beijing-based investment bank Chanson & Co.
While Chen is still slightly behind the net worth of Nvidia founder Jensen Huang, he is the world’s third-richest person aged 40 and under, according to the index, behind Lukas Walton and Mark Mateschitz, heirs to the Walmart and Red Bull fortunes, respectively.
Cambricon’s shares, and by extension Chen’s net worth, soared in August as Beijing urged local companies to avoid using market leader Nvidia Corp.’s H20 processors, especially for government-related purposes.
The company stepped in to calm investor frenzy around its shares by highlighting the difficulties of climbing the tech ladder by stating in a filing to the Shanghai Stock Exchange in August that it was still operating under US sanctions. It also eliminated speculation about non-existent products in the pipeline.
Broker notes also mention the upcoming Siyuan 690 chip around the same time, but that chip is believed to be at least a few years behind Nvidia’s related product.
“It’s too early to tell whether Cambricon or Huawei, China’s leading AI chip designer, will become China’s Nvidia, because it’s extraordinarily difficult to quickly copy Nvidia’s entire stack, including its CUDA ecosystem,” said Sunny Cheung, a researcher at Washington-based think tank Jamestown Foundation, referring to the proprietary programming language that ships with the AI chip giant’s hardware.
Cambricon did not respond to Bloomberg’s requests for comment.
Despite questions about Cambricon’s valuation, Chen’s path to success has become a case study for China’s state-backed academic pipeline and also spurred the surprise breakthrough of artificial intelligence startup DeepSeek and its millennial founder Liang Wenfeng.
Born in the southeastern city of Nanchang in 1985 to an electrical engineer father and a history teacher mother, Chen’s sharp intelligence was recognized early. He and his older brother, Chen Yunji, were quickly recruited into a program for gifted students at Hefei’s elite University of Science and Technology of China, where he earned his doctorate in computer science in 2010.
From there, Chen joined his brother as a researcher at the computer institute of the Chinese Academy of Sciences, the center of the country’s scientific ambitions, funded by state coffers.
The brothers first gained wider attention at this point with their internationally recognized academic paper on DianNao accelerators in 2014. A year later, they released their first chip, a brain-inspired processor for deep learning. This component was named Cambricon, named after the Cambrian explosion to indicate an early evolutionary starting point for Artificial Intelligence.
In 2016, the Cambricon project was canceled and established as a company, with the academy as the first financial supporter.
Huawei made its first breakthrough in 2017 when it used Cambricon’s AI processor technology to improve the photography and gaming capabilities of its Mate 10 smartphone. This partnership ended in 2019 when Huawei began developing similar technology on its own. Since then, Cambricon has gradually shifted its focus to designing and selling AI chips for both cloud servers and edge devices.
It was listed on the Science Technology Innovation Board in Shanghai in 2020, but was consistently in the red since its IPO before it started making a quarterly profit for the first time in the three months to December 2024.
It suffered a setback in 2022 when the U.S. Department of Commerce added Cambricon to a list of so-called entities that limit the company’s ability to access advanced Western technologies for its efforts to “acquire products of U.S. origin to support China’s military modernization.”
But US restrictions did little to hinder Cambricon’s hopes. When Washington expanded export controls to prevent Nvidia and AMD from selling high-performance AI chips to China, it created a supply gap. Beijing responded forcefully, requiring domestic technology firms to “buy local”; This means that Chinese companies must now source at least some of their chips from domestic manufacturers such as Huawei or Cambricon.
Demand has exploded. Cambricon’s revenue has increased by more than 500% in the last 12 months, even as it competes with Huawei and many other domestic startups.
“These spikes are driven directly by the urgent need for countries to access hardware infrastructure,” said Shuman Ghosemajumder, co-founder and CEO of Reken, a San Francisco-based artificial intelligence startup. “Similar to Nvidia, I think they will likely face a lot of change in stock prices once people decide exactly how much infrastructure is needed for practically useful generative AI models and how overstated those expectations are.”
–With help from Pui Gwen Yeung and Debby Wu.
More stories like this available Bloomberg.com
