Cambricon Becomes China’s First Trillion-Yuan AI Chip Stock as Wall Street Initiates Coverage

On June 30, Cambricon [SHA: 688256] became the first company listed on Shanghai’s STAR Market to cross a CNY1 trillion market capitalization, closing at CNY1,595.55 per share, a 7.7% single-day gain that capped a 71% year-to-date rally. The milestone arrived just as Macquarie analyst Eugene Hsiao initiated coverage on five Chinese AI chip companies, Cambricon, Huawei, Hygon, Biren, and MetaX, calling this the best entry point the sector has offered. Cambricon was named Macquarie’s top pick, with Hsiao projecting more than 50% upside from current levels.

The convergence of a landmark valuation event and the first major Western brokerage coverage of the sector in years marks a turning point for China’s domestic AI semiconductor industry. For most of the past three years, Chinese AI chip companies have operated largely out of view of international institutional investors, constrained by US export controls, limited English-language disclosure, and the shadow of Huawei’s ongoing sanctions battles. Macquarie’s initiation signals that the sector has matured enough, in revenue, product adoption, and policy visibility, to warrant dedicated analyst coverage.

(Related: China’s Domestic AI Chips Have Entered the Training Stack — and the Efficiency Gap Is Closing)

Why Cambricon Leads the Pack

Cambricon’s path to CNY1 trillion is rooted in its position as the most commercially focused pure-play AI chip designer in China. Unlike Huawei, which treats chips as one component of a vertically integrated stack, or Biren and MetaX, which are earlier-stage, Cambricon has been shipping AI inference and training chips at scale for several years. Its MLU series of AI accelerators are deployed across cloud platforms, smart city infrastructure, and edge devices, a breadth of deployment that gives it revenue visibility that peers cannot yet match.

Morgan Stanley’s projections reinforce the bull case: China’s AI chip market is forecast to grow from $21 billion in 2025 to $67 billion by 2030, a more than threefold expansion driven by surging demand for AI workloads in cloud computing, autonomous vehicles, and industrial automation. In that context, Cambricon’s current valuation, even at CNY1 trillion, implies a market share assumption that is far from aggressive, given its existing customer base and the structural tailwinds behind domestic chip adoption.

The government’s “secure and reliable” certification program has also become a meaningful commercial catalyst. In May 2026, nine AI chips from Huawei, Alibaba, and MetaX received the designation, a government-backed validation that signals to state-owned enterprises, banks, and telecoms that these chips meet national security and reliability standards. For Cambricon, the certification ecosystem creates a procurement preference that is difficult for foreign competitors to overcome, regardless of performance benchmarks.

(Related: Huawei and China Mobile Validate China’s First Carrier-Grade AI Inference System on Domestic Chips)

The Broader Sector Picture: Five Companies, One Thesis

Macquarie’s decision to initiate coverage on all five companies simultaneously, rather than singling out one, reflects a thesis about the sector as a whole rather than any single company’s idiosyncratic story. Huawei’s Ascend chips remain the dominant force in high-performance AI training, and the company’s plans to launch Ascend products in South Korea in Q4 2026 suggest it is beginning to test international markets. Hygon, which produces x86-compatible processors with AI acceleration features, serves a different segment of the market, enterprise servers and government infrastructure, where its compatibility with existing software stacks is a competitive advantage.

Biren and MetaX represent the higher-risk, higher-upside end of the coverage universe. Both are developing chips targeting the data center AI training market directly, competing with Nvidia’s H100 and H800 equivalents. Neither has reached Cambricon’s scale of commercial deployment, but both have received government support and are embedded in the national AI infrastructure buildout that Beijing has been accelerating since 2024.

The common thread across all five is the structural shift in China’s procurement behavior. As US export controls have tightened access to Nvidia’s most advanced chips, Chinese cloud providers, model labs, and enterprise buyers have been forced to qualify domestic alternatives. That qualification process — which involves months of integration testing, software optimization, and performance benchmarking, is now largely complete for the leading domestic chips. The result is a captive, growing market insulated from foreign competition by policy, not just preference.

What the Trillion-Yuan Milestone Means for Investors

The STAR Market, launched in 2019 specifically to support high-technology companies, has produced several large-cap success stories, but Cambricon’s crossing of the CNY1 trillion threshold is qualitatively different from earlier milestones. It places Cambricon in the same valuation tier as some of the largest semiconductor companies in the world, and it does so at a moment when the company’s revenue base, while growing rapidly, is still a fraction of what its market cap implies about future earnings.

That gap between current revenue and implied future value is precisely what Macquarie’s initiation is designed to help institutional investors navigate. By providing a structured framework for thinking about China’s AI chip market size, competitive dynamics, and policy tailwinds, Macquarie is enabling a class of investors who previously lacked the analytical infrastructure to take positions in the sector. The timing, with Cambricon’s stock already up 71% year-to-date, suggests that Macquarie sees further upside even after the rally, rather than initiating at a peak.