The firm’s shares have skyrocketed in recent years, propelling Tianshi, who owns roughly 29% of the company, into the ranks of China’s wealthiest, with a net worth of US$24.1 billion as of late October, according to Fortune.
![]() |
|
Chen Tianshi, founder and CEO of AI chip startup Cambricon Technologies, delivers a speech at Zhongguancun International Innovation Center on March 30, 2025 in Beijing, China. Photo by VCG via Reuters |
The surge has been fueled by expectations that the firm could become a serious challenger to U.S. chip giant Nvidia in the mainland market, even though it was launched less than a decade ago.
Cambricon was established back in 2016 by Yunji and Tianshi, who are now 42 and 40, respectively. Sons of an electrical engineer and a history teacher in Nanchang, the capital of Jiangxi province, they both went on to work at the Chinese Academy of Sciences (CAS), China’s leading science research institution.
There, the brothers studied in the elite "genius youth class," which was designed to train top scientific talent, and later joined the Institute of Computing Technology. Yunji specialized in semiconductors while Tianshi focused on artificial intelligence, as reported by the South China Morning Post.
Around 2010, they started looking into the idea of using chips to accelerate deep learning algorithms, or "AI plus chip." Their work gained recognition in 2014, when a co-authored paper on machine learning won the top award at ASPLOS, a leading computer architecture conference.
Heading a team of 20 in a 30-square-meter lab, the brothers then dedicated themselves to work and, in 2015, created a "deep learning dedicated processor prototype chip." They named it Cambricon, based on the Cambrian era of rapid evolutionary expansion.
With backing from CAS, the fledging group officially spun off in 2016 to launch Cambricon Technologies, with Tianshi as chairman and CEO and Yunji as chief scientist. The firm’s first major customer was tech giant Huawei.
Cambricon went public on Shanghai’s STAR Market in 2020, with shares jumping 230% on its debut. Its market capitalization surpassed 100 billion yuan within just four years.
The firm has also turned profitable. It posted its first quarterly profit, of 272 million yuan, in the fourth quarter of 2024 as revenues climbed 75.5% year-on-year to 989 million yuan. By the third quarter of 2025, its sales had jumped to 1.73 billion yuan while net profit for the quarter hit 567 million yuan, as against the 194 million yuan loss a year earlier, per Bloomberg.
Cambricon in August said it expects full-year revenues to reach 5-7 billion yuan, up from the previous year’s 1.2 billion yuan.
Looking further ahead, Goldman Sachs forecasts the company’s revenue could reach 13.8 billion yuan next year and its share of China’s AI chip could rise from 3% to 11% by 2028.
![]() |
|
A photo of Cambricon founders Chen Tianshi (left) and Chen Yunji. Photo from X |
Yet the brothers’ journey was far from easy. Yunji has shared that a disagreement over a research issue once escalated into a physical fight between him and his brother. Nonetheless, he said they both have a passion for chip design and have shared both "laughter and tears" while developing deep-learning processors, or AI accelerators.
"As brothers, we can put aside all pretence and honestly face all kinds of problems," Yunji told official media China Science Daily in an interview last September. "Perhaps that’s the reason we can rapidly advance our cross-disciplinary research into chip design and AI algorithms."
The elder Chen has also shared that back when they first envisioned AI chips, the technology had yet to gain traction and their idea was often brushed off as a "wild fantasy." It faced skepticism from the industry and failed to secure a 200,000 yuan research grant.
"It was like fumbling in the dark. No one knows which way was right, where the exit was, or even if there was an exit at all," Yunji recalled.
Another major setback came in 2019, when Huawei stopped using Cambricon’s technology for its smartphones, cutting off most of the firm’s business. Huawei had accounted for 98% of Cambricon’s revenues the year before that.
In response, Cambricon made a bold pivot that put it directly in competition with its former client: it shifted its focus to cloud computing accelerators, which handle complex AI tasks.
Six years later, the move has positioned Cambricon as Huawei’s main challenger in China, with investors betting that the former could emerge as a key beneficiary of Beijing’s push for greater technological self-sufficiency and reduced reliance on U.S. chips from Nvidia.
Several industry insiders, including a ByteDance AI engineer, noted that Cambricon’s software compatibility made its products easier to use than Huawei’s Ascend chips.
"Cambricon struggled to gain traction until the end of 2024, when it collaborated with ByteDance to make its chip more compatible with algorithms trained on Nvidia’s ecosystem," Lin Qingyuan, a semiconductor analyst at equity research and brokerage firm Bernstein, told the Financial Times.
Many of China’s top AI chip buyers, such as China Telecom, Alibaba, Tencent and Baidu, also rival Huawei in fields like network equipment, cloud services and self-driving vehicles. Hence, there is a strong incentive for them to support alternatives to the tech giant.
Cambricon has faced supply challenges after selling the bulk of its first-quarter chips to ByteDance, according to sources familiar with the matter. Its immediate priority is expanding production with its manufacturing partner, Semiconductor Manufacturing International Corporation, to meet the strong demand.
"Our customers have a very high rating for its products after testing," a salesperson at Inspur, one of the biggest data center assemblers in China, said. "There just isn’t enough supply."
Other challenges lie ahead. Cambricon was added to the U.S. Entity List in 2022, limiting its access to American technology and advanced manufacturing processes.
In August, the firm issued a risk warning to investors, noting that its share price had far outpaced industry peers and that its valuation might not align with its current fundamentals.
Analysts have also warned that the company will remain a smaller player compared with Huawei as the latter continues to address software issues that have hindered its chips. Huawei is currently the only chipmaker offering a viable alternative to Nvidia’s NV-Link, and this could give it an advantage in AI training chips.
Still, some investors believe Cambricon could one day become China’s answer to Nvidia. Its co-founder remains determined as well.
The company "must shoulder the mission and responsibility of our era" to advance AI technology, Tianshi said during an investor call earlier this year.