Li Auto, BYD, Nio, and Xpeng have all unveiled custom silicon for autonomous driving in the past year, turning chip design into the latest front in China’s EV tech war.
China’s four largest EV makers have all unveiled custom autonomous-driving chips in the past year, aiming to cut costs and reduce reliance on Nvidia and Horizon Robotics. The shift is driven by US export controls and the rapid commoditisation of smart-driving features in China’s mass market.TL;DR
A new arms race is under way in China’s electric vehicle industry, and it is not about batteries, range, or price. It is about who controls the silicon that makes cars drive themselves.
In the past 12 months, four of China’s largest carmakers have unveiled proprietary chips designed specifically for autonomous driving, each aiming to reduce dependence on Nvidia and Hong Kong-listed Horizon Robotics, the two dominant suppliers of smart-driving computing hardware in the Chinese market.
The latest entry came on Monday, when Li Auto unveiled the Mach M100, a 5-nanometre chip tailored for its new L9 Livis SUV. A single M100 delivers 1,280 trillion operations per second (TOPS), a measure of how fast an AI processor can handle incoming sensor data, with an 82% utilisation rate.
The 💜 of EU tech
The latest rumblings from the EU tech scene, a story from our wise ol' founder Boris, and some questionable AI art. It's free, every week, in your inbox. Sign up now!
Weeks earlier, BYD debuted the Xuanji A3, a 4nm chip already in mass production. Three Xuanji A3 chips working together deliver more than 2,100 TOPS, enough to support Level 3 and Level 4 autonomous driving.
Nio has been shipping its own 5nm NX9031 chip since the ET9 sedan launched in early 2025, and has since deployed it across its entire main-brand lineup and into its mass-market Onvo sub-brand. Xpeng, meanwhile, is rolling out its proprietary Turing chip, which the company says delivers up to 2,250 TOPS per unit.
Two forces are pushing Chinese carmakers to design their own silicon. The first is geopolitical.
US export controls have steadily tightened access to Nvidia’s most powerful chips, and the threat of further restrictions has made self-reliance a strategic imperative.
The second is economic. Smart-driving features have gone from a luxury-segment perk to a mass-market expectation in under a year.
The penetration rate of passenger cars equipped with driver-assistance systems reached 67.6% in China in 2025, according to Horizon Robotics’ annual report. The share of vehicles fitted with mid-to-high-level systems, including urban navigate-on-autopilot, nearly doubled to 42.6% from 21.6% the year before.
The shift has been especially dramatic in the cheapest cars. In the segment below 200,000 yuan ($29,545), which accounts for roughly 65% of China’s passenger-car sales, the penetration rate of advanced driver-assistance systems rose from 5% at the start of 2025 to more than 50% by year-end.
That pace of adoption has created a brutal cost problem. BYD’s God’s Eye B system, which includes urban NOA, is now available across all its models for 12,000 yuan.
Goldman Sachs analysts said the package saw a take-up rate of more than 60% on models including the entry-level Seagull, effectively bringing the floor price of a BYD with urban NOA to 78,800 yuan.
Goldman noted that Nvidia and Horizon Robotics posted gross margins of 75% and 65% respectively in 2025. For carmakers building millions of vehicles a year, those margins represent a significant cost they believe they can claw back by designing their own chips.
Huatai Securities estimated that Nio’s switch to its in-house NX9031 chip cut per-vehicle costs by 10,000 yuan. China’s broader pivot toward application-specific silicon, driven partly by export controls and partly by economics, is accelerating across the AI industry.
For carmakers, the advantages of custom chips go beyond unit costs. Tighter integration between hardware and software, faster iteration, and greater control over product roadmaps all matter in a market where autonomous-driving capability is rapidly becoming the primary differentiator.
Nvidia remains dominant globally, and its Drive Thor platform is the benchmark for next-generation autonomous driving compute. But in China, its largest automotive market outside the US, the ground is shifting.
BYD, the world’s largest EV maker by volume, no longer needs Nvidia for its smart-driving stack. Nio has replaced Nvidia silicon across its entire lineup.
The competitive intensity in Chinese EVs now extends from batteries and pricing into the chip architectures that power autonomy.
Jensen Huang has warned that Chinese companies running advanced AI on non-Nvidia hardware would be a “horrible outcome” for America. In China’s autonomous-driving sector, that outcome is already arriving.
Get the most important tech news in your inbox each week.
