Market Context 📈

Cambricon’s stock has skyrocketed, fueled by a jaw-dropping revenue surge and a shift to profitability, transforming sentiment around China’s AI hardware supply chain. The company’s trailing P/E ratio has soared past 4,000, briefly surpassing Kweichow Moutai to make Cambricon China’s priciest stock by this metric. This milestone reflects the booming enthusiasm for domestic AI accelerators, though it comes with execution risks in a sanction-constrained environment. China’s push to localize AI compute and reduce reliance on foreign hardware has created a perfect storm for companies like Cambricon, with local demand tied to ecosystems like DeepSeek driving momentum.

Earnings Surge 💰

In the first half of 2025, Cambricon reported revenue of RMB 2.88 billion, a staggering 4,000%+ increase year over year, and a net profit of RMB 1.04 billion, reversing a RMB 0.53 billion loss from the same period in 2024. This marks three consecutive profitable quarters since late 2024. While Cambricon’s revenue is modest compared to Nvidia’s, the growth rate and profitability shift signal strong adoption among Chinese AI model developers and internet platforms. Management credits “continued market expansion” and support for AI application deployment, aligning with China’s localization trends.

Valuation and Moutai Comparison 📊

Cambricon’s share price has more than doubled month over month at times, pushing its market capitalization toward RMB 580 billion and its trailing P/E above 4,500. This briefly outstripped Kweichow Moutai, signaling a shift in China’s A-share market where AI compute now overshadows consumer staples. However, the lofty valuation raises questions about earnings durability versus momentum-driven multiples. Analysts point to regulatory support, domestic demand, and model optimizations for local chips as key drivers, but sustainability remains a concern.

Product and Ecosystem 🛠️

Cambricon specializes in AI accelerators for cloud, edge, and data center clusters, with its Siyuan roadmap designed to compete within domestic constraints. Reports highlight growing compatibility with Chinese AI frameworks and training approaches optimized for home-grown silicon, amplified by buzz from model developers like DeepSeek. This ecosystem alignment is critical for sustaining orders through 2025 as Chinese internet giants and AI startups scale deployments on sanction-resilient hardware.

Case Study: DeepSeek-Driven Demand 🌟

A clear driver of Cambricon’s momentum is DeepSeek’s disclosure that its latest model was trained using a new data format optimized for domestic chips. This sparked investor and customer interest, boosting Cambricon’s stock in August 2025. Industry reports confirm Cambricon’s compatibility with this ecosystem, reinforcing procurement confidence. As a result, Chinese model developers and internet platforms are increasingly trialing and scaling on Cambricon’s accelerators to secure supply and cost visibility, fueling the H1 2025 revenue surge.

Risks and Sustainability ⚠️

Despite the explosive growth, Cambricon’s future hinges on scaling production, competing with other Chinese chipmakers, and maturing software stacks for domestic accelerators. The sky-high valuation assumes flawless execution and continued policy support. Any issues with yields, ecosystem readiness, or procurement cycles could trigger volatility. Compared to global giants like Nvidia, Cambricon’s revenue base is still small, making quarter-to-quarter fluctuations a key risk to watch.

Investment Takeaway 💡

Cambricon’s 4,000%+ revenue surge, profitability swing, and ecosystem pull from domestic AI leaders have made it a poster child for China’s AI chip boom. Its valuation reflects both current momentum and high expectations for future scale-up. For investors, the focus is shifting from survival to sustainability: pipeline visibility, product cadence, and software maturity will determine whether Cambricon can justify its “priciest stock” status or face a re-rating as competitive and execution risks emerge. Near-term policy support and localization trends remain strong, but long-term success depends on multi-node roadmaps and broader customer adoption across cloud and edge.

FAQs ❓