AXERA annual revenue of 560 million: explosive growth in smart cars and edge AI inference business

Leidi Network, Letian, March 27 AXERA (0600.HK) today released its performance report for the fiscal year 2025. In 2025, the company achieved total revenue of 562 million yuan, a year-on-year increase of 18.8%; gross profit reached 121 million yuan, with a gross margin stable at 21.6%. The proportion of emerging businesses in overall revenue rapidly increased, with the smart automotive business achieving a key breakthrough, generating annual revenue of 48.17 million yuan, a year-on-year increase of 618.2%; the edge AI inference business generated revenue of 43.6 million yuan, a year-on-year increase of 134.6%. Smart Automotive Business Moves Towards High-End Against the backdrop of the wave of automotive intelligence and accelerated integration of the industrial chain, AXERA achieved a milestone breakthrough in 2025, with the annual shipment of its smart automotive SoC exceeding 630,000 units. Currently, the M55H chip has entered mass production in several major manufacturers and has been deployed in practice. The SoC M57, designed for the global market, has become the selected platform for several internationally renowned Tier 1 suppliers and will be mass-produced for overseas markets. The high-end assisted driving SoC M97 chip was taped out in October 2025 and lit up after the return of the chip in February 2026. This chip has a computing power exceeding 700 TOPS, supports "Urban NOA," and can fully cover the intelligent driving requirements from L2+ to L3/L4 levels, making it the core product for the company's move towards high-end smart automotive business. In terms of market expansion, AXERA collaborated with international Tier 1 suppliers to achieve mass production of L2 ADAS products and secured overseas project designations in 2025. In 2025, more than 25 new designated projects were added, with over 15 collaborating OEM brands, including more than 5 international brands. In November of the same year, the company led the establishment of a joint venture, Chongqing Chuangyuan Zhihang, focusing on the research and sales of high-end assisted driving chips, with the company holding the majority of seats on the board as the single largest shareholder, thus having control. Edge AI Inference Business Expansion Strategy Deepens The edge AI inference business is another highlight of the company's forward-looking layout, achieving revenue of 43.6 million yuan in 2025, a significant year-on-year increase of 134.6%, realizing a deep strategic expansion. At the beginning of 2025, breakthroughs in large models spurred demand for local real-time inference, low latency, and high privacy protection on the edge, while effectively reducing cloud token costs. AXERA's AX8850N chip, with its advanced architecture and computing power advantages, as well as its rare native support for Transformers, achieved adaptation for large models such as DeepSeek R1 Distill and Qwen 2.5 on the edge This technological breakthrough helps customers successfully bring large model computing power down to the device level, enabling AI Agents to think and make decisions locally. Based on this, AXERA's AI computing power solutions have successfully penetrated emerging sectors such as industrial cameras, drones, mobile robots, and AI Boxes. Notably, based on the OpenClaw platform, the company's smart terminal product AXClaw Box provides a local AI computing power solution that saves tokens, is secure and reliable, and offers high cost-performance for smart home and office scenarios. The foundation of terminal computing continues to solidify, with delivery scale reaching new highs. As the cornerstone business of the company, the terminal computing sector further solidified its leading position in the global visual terminal chip market during the reporting period. Relying on its self-developed AXERA AI-ISP technology, the company's "Black Light" series (including AX630C, AX620Q, etc.) SoC chips saw a significant year-on-year increase in sales, fully covering mainstream application scenarios such as long-range, battery-powered, multi-camera, and AOV. At the same time, this technological advantage has strongly supported the overseas expansion strategy of domestic terminal brands, rapidly promoting the series products in overseas markets. In terms of product iteration, the newly released AX615 series integrates the next-generation AI-ISP and video encoder, prompting several leading customers to initiate project approvals, laying the foundation for the company's continued growth in this field by 2026. R&D expenditure of 596 million in 2025. AXERA always regards technological innovation as the lifeline of corporate development. In 2025, the company will invest 596 million in R&D, focusing on perception and computing, self-developing core technologies of "AI-ISP + NPU," and building a unified technology platform. Stable R&D investment, significantly improved R&D efficiency, effectively supports the simultaneous development of multiple high-end SoCs. In addition, the company's technological moat is becoming increasingly robust, with 81 new intellectual property applications and 69 new authorized intellectual properties during the reporting period; the cumulative number of intellectual property applications reached 643, with 317 authorized, further strengthening the technological barriers. AXERA stated that in 2025, adhering to the technical strategy of "universal technology, specialized chips, and ecological synergy," while consolidating its global leadership position in visual terminal chips, it will build the smart automotive business into a strong second growth engine, and strategically layout edge AI for the future, forming a business layout of "scalability, high-end, and globalization." Looking ahead to 2026, solutions driven by the next-generation edge and edge AI chips are expected to be more widely adopted in smart vehicles, edge AI, and other massive physical devices. AXERA stated it will continue to increase R&D investment, focusing on the continuous improvement of computing power, energy efficiency, and software usability, to meet the extreme demands of next-generation AI applications for underlying capabilities with higher performance computing platforms. At the same time, the company will seize the era's opportunities for the deep integration of artificial intelligence and the real economy, promoting the inclusiveness of intelligent computing and injecting lasting momentum for long-term, sustainable value creation ———————————————