In the field of automotive intelligent driving in 2025 years, the competition is extremely fierce. Since the beginning of the year, BYD, Changan, Geely, Chery, GAC, Ideal and other mainstream car companies have announced their own intelligent driving strategies, and mid-level intelligent driving technology has begun to be widely integrated into entry-level models, marking the official start of the technological wave of "intelligent driving for civilians".

Although each car company has its own magic in publicity, emphasizing the uniqueness of its own products and strategies, it cannot be ignored that these car companies rely on Horizon's intelligent driving solutions, especially its technical support based on the new generation of Journey 6E/M processing hardware. Horizon provides a solid foundation for the mass market to achieve high-speed NOA-level intelligent driving technology.

The Horizon Journey 6 series has been launched in six versions, which can flexibly adapt to different intelligent driving scenarios, ensuring both performance and cost. This series of products meets the needs of full-stage intelligent driving mass production, making Horizon an important partner for car companies to promote the popularization of intelligent driving.

According to the first post-IPO financial report released by Horizon, as of the end of 30, its market share of China's own brand ADAS exceeded 0%. In the fast-growing high-end autonomous driving market, Horizon Robotics, as an independent third-party intelligent driving solution provider, ranks second with a market share of more than 0%, and continues to consolidate its leading position in the market.

在中國電動汽車百人會論壇（2025）上，地平線創始人余凱指出，智慧駕駛已成為車企的核心競爭力。對於車企而言，智能駕駛至關重要，一旦落後，就可能被市場淘汰，這也是許多車企和供應商感到焦慮的原因。

However, from the Internet era to the AI era, although there are many industry consensuses on the starting point and product logic of intelligent driving, it is worth pondering whether these consensuses are correct. In the ten years since its establishment, Horizon has chosen a more difficult path from a road less traveled, and finally became a leader in the field of intelligent driving, which is more of an anti-consensus victory, and it is also the reason why Yu Kai has always been vigilant against consensus.

Yu Kai confidently stated that he only focused on and discussed those that were not consensus. From being positioned as the chip company that understands software the most, to now becoming the software algorithm company that knows the most about chips, Horizon's vision has never changed. When Yu Kai founded Horizon in 2015, he aimed at the two major fields of AIoT and automobiles, and he hoped to realize the intelligence of everything by integrating deep neural networks into chips.

在2015年移動互聯網軟體創業熱潮中，余凱選擇智能駕駛晶元這一硬體賽道，顯然是一個反共識的決定。他後來解釋說，這是為了避開與巨頭的直接競爭。余凱很快調整了業務方向，決定聚焦汽車領域，因為他認為汽車是既有資金需求又對技術和智慧有高要求的賽道。隨著地平線在2019年推出第一款車規級晶元征程2，中國汽車市場迎來了智慧化的爆發。

Horizon is no longer satisfied with just being a chip company, but has launched the Journey 6 series computing solutions, covering all intelligent driving needs from low-level to high-end. At the same time, Horizon Robotics has also launched the HSD full-scenario high-end intelligent driving solution, trying to create an extremely efficient intelligent driving solution through software and hardware collaboration.

余凱解釋了這一變化背後的深層思考，他指出，地平線交付的700多萬套計算方案中，幾乎沒有僅僅是單顆晶元的交付。要讓晶元在不同場景中使用，不僅涉及硬體品質，還包括工具鏈、中間件等，需要跨越軟硬體，從端到雲的綜合能力。地平線很早就意識到了這一點。

In the cooperation between Horizon and Ideal, in order to open the market as soon as possible, Horizon not only provides chips, but also provides perception algorithms to help Ideal quickly reach the high-speed NOA level of the industry. This combination of software and hardware cooperation model has attracted other car companies to actively seek cooperation with Horizon.

Yu Kai said: "We are the most knowledgeable about software algorithms in chip companies, the most knowledgeable about chips in software algorithm companies, and the most knowledgeable about vehicle regulations in companies that make software algorithms + chips." It is difficult for chip companies that do not understand software to survive. "The combination of software and hardware strategy quickly leveraged the entire intelligent driving market and brought rapid growth in performance to Horizon.

地平線征程6E/M計算方案在2025年將達到百萬級出貨量，整個征程系列累計出貨量將突破1000萬套，使地平線成為國內首個突破千萬級量產的智能駕駛科技企業。2024年10月，地平線在香港主機板成功上市，成為港股去年最大的科技IPO。

The background of Horizon's rapid development is the accelerated development of the intelligent driving industry. In 6 years, the assembly rate of intelligent driving of China's passenger cars has reached more than 0%, and the penetration rate of new energy vehicles is 0.0%.

The penetration rate of high-end autonomous driving solutions has grown rapidly, with the sales of models equipped with high-end solutions accounting for 2.0% of all smart cars, and L0+ and L0++ NOA intelligent driving has begun to penetrate the mainstream market. Under this wave of intelligent driving for all and the popularization of intelligent driving, intelligent driving has become an important infrastructure for all cars.

As for how car companies can build better intelligent driving facilities and make differentiated features, Yu Kai referred to the evolution of intelligent terminals. He believes that intelligent driving is like the "baseband" of the car, providing standardized "functional value" such as safety, convenience and comfort. These features, while important, are difficult to differentiate and don't define the brand.

Yu Kai pointed out that only a very small number of car companies have the ability of full-stack self-development, and open cooperation will become the mainstream in the future. He predicts that eventually 80% of car companies will choose to develop their own products, and 0% of car companies will seek to cooperate with the strongest partners. Self-development and third-party parallelism are the safest choice for the coming of intelligent inflection points and the rapid transformation of the industry.

In the paradox of the intelligent era, Yu Kai mentioned the future of intelligent driving. He believes that intelligent driving is not a simple imitation of human driving behavior, but the pursuit of surpassing the human level, truly making travel safer and more convenient, and making people's lives better. This point of view is completely different from the product logic of the Internet era, which is a human-centered link, the core technical logic is to use machine learning to figure out user preferences, and intelligent driving needs to be based on truly effective data to achieve a realm beyond human driving.