Huawei's Qiankun ADS R&D Budget Raised to Over 19 Billion Yuan, Jin Yuzhi Says

Jin said China's new energy vehicle industry has grown over a decade into the world's largest market, while domestic competition has become increasingly intense and companies are examining how to achieve sustainable operations. Against that backdrop, he said, automakers and suppliers should pursue healthy industrial division of labor and coordination.

He noted that many automakers are considering in-house development of chips, intelligent driving systems and smart cockpit systems. Using chips as an example, Jin said developing a chip is not difficult, but once a company enters the field it must complete iterative upgrades every two to three years; without sufficient market scale, the large upfront investment is difficult to amortize.

In intelligent driving, Jin said safety is always the first priority for assisted driving. Huawei Qiankun has placed safety at the highest priority since entering the field, he said. Cumulative assisted driving mileage has reached 15.7 billion kilometers, and when ADS is activated, driving safety is 5.5 times that of human driving, according to data that Jin said is publicly available on the company's website.

He said the company discloses the data transparently to make the point that safety is a non-negotiable precondition when using assisted driving systems. Facing competitive pressure, many companies first consider cutting costs and prioritize full in-house development, but Jin said the investment account must be calculated clearly: building a mature, usable autonomous driving system is not a one-time effort after spending several billion yuan.

Huawei Qiankun's R&D budget for this year has been raised from 18 billion yuan at the beginning of the year to more than 19 billion yuan, Jin said. Even if an automaker can sell 1 million vehicles a year, a scale many automakers do not reach, it would still have to bear high annual costs at the software level, while investment in manpower and computing power for model training cannot be interrupted, he said. This year's computing power expenses reached 7 billion to 8 billion yuan, according to Jin.

Once a company enters this track, there is no turning back, he said. Considering the full costs of hardware, domain controllers and software iteration, the threshold for full in-house development is very high and is not necessarily the most efficient path for most automakers, Jin said.

He said Huawei has consistently pursued mutually beneficial cooperation with automakers. The company offers both high-end solutions for premium models and cost-effective assisted driving solutions for a broader market that also uphold safety bottom lines, according to Jin. Through specialized division of labor, automakers can focus on vehicle definition, chassis, vehicle integration and user operations, while suppliers apply their own technological accumulation, he said, adding that each side should perform its role to improve product experience, enhance the competitiveness of the entire industry and ultimately benefit users.