Cloudwalk on the Eve of Explosion: AI Chip Track, Cambrian-Style Growth Repeating

Recently, Cloud Tianli Fei officially won the bid for the Zhanjiang City AI Penetration Support New Quality Productivity Infrastructure Construction Project, with a bid amount of 420 million yuan. The project will center around the company’s self-developed AI inference chips to create a domestically produced trillion-calorie-level inference cluster, implementing the “National Model, National Chip” integrated intelligent computing platform. This marks a critical step for Cloud Tianli Fei’s self-developed chips from technological validation to large-scale commercial use.

From breakthroughs in self-developed chips to the implementation of major projects, from technical deepening to commercialization, Cloud Tianli Fei is rapidly progressing along a growth path highly aligned with Cambrian’s, with a new wave of Cambrian-style growth in the AI chip sector reenacting.

/ 01 /

Industry Rule: AI Chip Growth Depends on Order Fulfillment, Not Just Technical Strength

There is a simple and brutal growth rule in the AI chip industry: large projects are densely implemented, directly leading to vertical revenue takeoff. This has been validated by Cambrian’s development history. The market used to obsess over chip parameters and technical concepts, but ultimately, financial reports are the most honest, and order scale is the most convincing.

Looking back at Cambrian’s growth curve, it’s a typical order-driven model: in 2023, revenue was only 700 million yuan; in 2024, driven by large orders, revenue rose to 1.174 billion yuan; in 2025, revenue exploded to 6.497 billion yuan, a year-on-year increase of 453%, achieving full-year profit after continuous losses in just one year. This demonstrates that the “growth” of AI chip companies is never a fleeting pilot project but a qualitative change characterized by billion-yuan orders becoming routine and business being replicable.

History may not simply repeat itself, but it often rhymes surprisingly. Currently, Cloud Tianli Fei’s development pace perfectly aligns with this mature growth logic, shifting from isolated technological breakthroughs to full-scale business growth. The sector’s wind is already in the sails.

/ 02 /

20 Billion Yuan in Orders as Foundation: Three Major Core Projects, Moving from Pilot to Large-Scale Deployment

Unlike most companies in the industry that rely on POC pilots and demo projects, Cloud Tianli Fei has secured real, long-term orders totaling over 2 billion yuan in just two years, with three core projects forming a solid business foundation.

Among them, the 1.6 billion yuan computing power order from Deyuan Fanghui directly secures a stable long-term revenue stream, strengthening the company’s fundamental development; the 130 million yuan Tianfu Intelligent Computing Center project helps it establish a regional computing hub position and opens channels for local government and enterprise cooperation.

Recently, Cloud Tianli Fei successfully won the bid for the Zhanjiang City AI Penetration Support New Quality Productivity Infrastructure Construction Project, with a bid amount of 420 million yuan. The project centers on the company’s self-developed AI inference chips to build a fully domestically produced trillion-calorie inference cluster, deeply adapted to domestic large models, and constructing the “National Model, National Chip” integrated intelligent computing base. This signifies that Cloud Tianli Fei’s self-developed chips have officially entered a new stage of large-scale commercial use and infrastructure-level deployment.

The Zhanjiang Intelligent Computing Center project carries the Tianli Fei self-developed chips and has created an application ecosystem for domestic chips in internet, software, and telecom customers. Its strategic value and business model closely mirror Cambrian’s key path from technological breakthrough to industry-wide volume, with a new Cambrian-style growth wave reenacting on Cloud Tianli Fei.

More importantly, Cloud Tianli Fei has already established a complete closed-loop chain of “self-developed AI chips + computing boards + intelligent computing centers + domestic large model adaptation.” The rapid success of the Zhanjiang Intelligent Computing Center relies on this full system of “domestic chips supporting domestic models and intelligent services for cities.” From chips to computing power, from computing power to scenarios, and from scenarios to ecosystems—each link is interconnected, and Cloud Tianli Fei now controls all of them.

/ 03 /

Low Base, High Flexibility: Repeating Growth Scripts to Capture New Inference Computing Power Opportunities

Compared to financial data, Cloud Tianli Fei is on the cusp of a key growth phase similar to Cambrian’s in 2024, with high growth potential hidden behind a low base. In 2023, Cloud Tianli Fei’s revenue was 506 million yuan; in 2024, it increased to 917 million yuan; and in 2025, it is expected to reach 1.308 billion yuan. This aligns closely with Cambrian’s revenue range before its explosive growth, indicating a golden period of accumulation before performance breakout.

Cambrian’s growth script is that after dense order fulfillment, revenue multiplies the following year; Cloud Tianli Fei not only replicates this logic but also upgrades through a reserve of over 2 billion yuan in secured orders. These projects are expected to convert into revenue within 1-2 years, fundamentally rewriting financial performance. Compared to Cambrian’s capture of the AI training boom, Cloud Tianli Fei is better positioned to seize the new industry trend—shifting focus from training to inference, with exploding inference computing demand, accelerated domestic substitution, widespread deployment of large models, and rapid construction of city intelligent computing centers. The certainty of its growth far exceeds industry averages.

The reshuffle in the AI chip sector is always an opportunity for practitioners who are willing to do the work. Cloud Tianli Fei is not simply copying the growth path of predecessors but has carved out its own growth curve through precise layout in inference, full-chain independence, and control. Its performance explosion is imminent.