Qingce GT7 with Huawei Ascend ADS 5: Autonomous Driving With No Network Onboard
On July 9, 2026, the Qingce GT7 began full-scale deliveries to showrooms. It is the first production car with Huawei's Ascend ADS 5 autonomous driving running directly onboard. There is no cloud computing and no latency from remote calls. The car drives at high speeds, in cities, in parking lots and on dirt roads, and it all works out of the box, without calibration for each route. RCA (Roaming Cruise Assist) changes lanes automatically without navigation. The parking system covers 120+ million parking spaces in China and runs onboard.
AI-processed from 36Kr (36氪); edited by Hamidun News
On July 9, 2026, Qingce (启境) began full-scale deliveries of the GT7, the first production car with Huawei's Ascend ADS 5 AI chip running directly in the vehicle. In practice, this means Huawei has managed to move autonomous driving from the cloud to the edge.
What comes in the box
The GT7 ships with two capabilities:
- RCA (Roaming Cruise Assist): route cruise control without navigation. The system sees the road, changes lanes on its own and follows the vehicle ahead. It works everywhere: at high speed, in the city and on dirt roads.
- Parking: the system covers 120+ million parking spaces across China, runs onboard and guides the car into a space without navigation.
All of this is native and works right after purchase. There are no map downloads, no calibration for local terrain and no cloud requests.
Why it matters
Until now, autonomous driving required either a constant cloud connection (with latency and dependence on 5G) or years of mapping work (every city needs its own map). Ascend ADS 5 runs locally onboard and adapts on the fly.
The Ascend 950, Huawei's new chip, is built for inference, meaning it runs finished models. The driving model is trained offline. Onboard, it simply runs: it reads camera frames and predicts the car's trajectory and the trajectories of other objects.
What it means
Huawei has shown that autonomous driving can run at the edge without the cloud. That shifts the competition: the contest used to be about data centers and the cloud, and now it sits at the boundary between the car and the network. The Qingce GT7 is not just a car. It is evidence that edge AI has moved from the lab into production.
- Nov 1, 2026New test catches hidden AI agent collusion by measuring how a partner reacts to a paraphrase
- Nov 1, 2026Illinois compiled 12 151 open problems and tested whether AI can solve them
- Oct 31, 2026AI engineer writes a single reward function for agent teams: win rate climbs from 51% to 90%
- Oct 31, 2026SkillMem stores an AI agent skill in 9.6 KB and speeds up a team 4.2x
Key Questions Answered
What comes in the box?
The GT7 ships with two capabilities: RCA (Roaming Cruise Assist): route cruise control without navigation. The system sees the road, changes lanes on its own and follows the vehicle ahead. It works everywhere: at high speed, in the city and on dirt roads. Parking: the system covers 120+ million parking spaces across China, runs onboard and guides the car into a space without navigation. All of this is native and works right after purchase. There are no map downloads, no calibration for local terrain and no cloud requests.
Why it matters?
Until now, autonomous driving required either a constant cloud connection (with latency and dependence on 5G) or years of mapping work (every city needs its own map). Ascend ADS 5 runs locally onboard and adapts on the fly. The Ascend 950, Huawei's new chip, is built for inference, meaning it runs finished models. The driving model is trained offline. Onboard, it simply runs: it reads camera frames and predicts the car's trajectory and the trajectories of other objects.
What it means?
Huawei has shown that autonomous driving can run at the edge without the cloud. That shifts the competition: the contest used to be about data centers and the cloud, and now it sits at the boundary between the car and the network. The Qingce GT7 is not just a car. It is evidence that edge AI has moved from the lab into production.
Delegate Deep Research, Coding & Multi-Step Workflows to Nomad Agent
Deploy autonomous task execution with Nomad Agent (agent.nomadnet.ai) — automated web research, data synthesis, code generation, and MCP tool orchestration.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.