8T8R 4D Millimeter-Wave Radar Debuts for Central Computing Vehicle Architectures

8T8R 4D millimeter-wave radar technology has taken a significant step forward as Huizhou Desay SV Automotive Co., Ltd. announced the launch of its first solution designed to support central computing satellite architectures. The company stated that the radar has already secured orders from major domestic automakers and is planned for mass production in 2026.

Higher accuracy and extended perception capability

The newly launched 8T8R 4D millimeter-wave radar adopts an 8T8R waveguide antenna solution, delivering around 30% higher angular accuracy compared with mainstream 6T8R configurations. This improvement enhances the radar’s ability to distinguish close-range objects in adjacent lanes, supporting more reliable perception in dense traffic scenarios.
In addition to accuracy gains, the radar achieves a point cloud density approximately 1.5 times higher than that of 6T8R solutions. Combined with the signal gain advantages of the waveguide antenna, weak target detection distance is extended by nearly 50%, effectively reducing false triggering of automatic emergency braking systems.

Robust performance in complex environments

Through multi-channel collaboration, the radar strengthens its anti-jamming capability and maintains stable perception in complex electromagnetic environments. This includes challenging operating conditions such as rain, fog, and tunnels, where signal interference and attenuation often affect conventional radar performance.
From a performance perspective, the product integrates high signal transmission efficiency and dense point cloud output within a compact form factor. The maximum detection range exceeds 300 meters, with the ability to identify traffic cones at 120 meters and stationary vehicles at distances of up to 280 meters.

Native support for central computing satellite architecture

Architecturally, the 8T8R 4D millimeter-wave radar natively supports satellite architectures, enabling direct transmission of high-fidelity ADC raw data to a central domain controller. This approach removes the need for traditional onboard edge processing units within each radar module.
By simplifying the radar hardware structure, the solution reduces per-unit hardware cost and power consumption while supporting sensor lightweighting and higher system-level integration. This centralized processing strategy aligns with the broader industry shift toward software-defined vehicles and centralized computing platforms, enabling scalable deployment of advanced driver assistance and autonomous driving functions.