Huawei to spend 80 billion yuan

Huawei said on April 23 it will spend 18 billion yuan on smart-driving research in 2026, including 10 billion yuan on computing power for training. (reuters.com) Senior vice president Jin Yuzhi said the company plans to spend 70 billion to 80 billion yuan on computing power over the next five years. He made the announcement at a Beijing event ahead of China’s biggest auto show. (reuters.com) (digitaltoday.co.kr) That money goes into the cloud systems that train driving software on huge volumes of road data before the software is pushed into cars. Huawei said the goal is to improve reliability for vehicles using its Qiankun advanced driver-assistance system and to win more automaker customers. (digitaltoday.co.kr) (reuters.com) Huawei’s Qiankun ADS platform had passed 10 billion kilometers of cumulative assisted-driving mileage by early April. The system is now installed in more than 50 models across 25 car brands, with 1.7 million vehicles delivered. (scmp.com) Huawei used the Beijing show to present itself less as a parts vendor and more as a branded automotive software platform. Reuters said Chinese buyers have helped make smart-driving features a selling point in the electric-vehicle market, especially in higher-end models. (reuters.com 1) (reuters.com 2) At the same event cycle, Huawei unveiled Qiankun ADS 5.0 and said the new system is aimed at scaling Level 3 autonomous-driving capability. Huawei also launched its Qiankun operating system for vehicles. (news18a.com) Richard Yu, who leads Huawei’s consumer and automotive efforts, told automakers the company can cut development costs, shorten testing cycles and speed model launches. That pitch is central to Huawei’s push to become the software and computing layer inside more Chinese cars. (digitaltoday.co.kr) (reuters.com) The spending plan shows how much of the self-driving race now depends on data centers as much as dashboards. Huawei is betting that more training compute will help keep Qiankun inside the next wave of Chinese vehicles. (reuters.com)