Denso Harnesses the Power of AMD Xilinx for Advanced LiDAR Platform
AMD’s adaptive computing tech is now powering Denso’s latest lidar platform, aiming to significantly improve the detection of pedestrians, vehicles, and more. This new lidar system promises to be over twenty times more precise, offering low latency for better resolution. Set to start shipping in 2025, it features AMD’s Xilinx Automotive (XA) Zynq UltraScale+ adaptive SoC, ensuring high safety standards with ISO 26262 Asil-B certification.
Denso is incorporating the AMD XA Zynq UltraScale+ multi-processor system-on-a-chip (MPSoC) into its single-photon avalanche diode (SPAD) lidar, renowned for delivering the highest point-cloud density in the market. Point-cloud density, which is akin to image resolution, ensures that more data points are captured for crucial decision-making in autonomous vehicles. Automakers prefer SPAD-based systems for their compact size, allowing them to fit more units in a single vehicle.
The adaptable XA Zynq UltraScale+ MPSoC helps Denso reduce the size of current lidar systems, enabling multiple lidars to work together for comprehensive forward and side views of a vehicle. This flexibility allows one device to handle multiple lidars, including future models, making designs more cost-effective and future-proof.
While current vehicles typically have one forward-facing lidar, future models will incorporate multiple systems for 360-degree coverage, essential for progressing from assisted driving to full autonomy. Beyond automotive applications, Denso’s lidar can be used in infrastructure monitoring, factory automation, and other areas.
Eiichi Kurokawa, head of sensing at Denso, expressed excitement about expanding their partnership with AMD, citing the high performance and scalability of AMD’s programmable silicon. The flexibility and safety standards met by the Zynq UltraScale+ MPSoC were key factors in their collaboration.
Denso’s SPAD lidar can generate over three million points per second at ten frames per second. Utilizing XA Zynq UltraScale+ MPSoCs for system monitoring ensures proper temperature and overall system functionality. The use of time-to-digital conversion optimizes system size and cost while maintaining high performance and data density.
Mark Wadlington, senior vice president at AMD, praised Denso’s precise lidar and highlighted the evolving tech requirements that necessitate improvements in sensitivity, density, and performance. AMD’s adaptive computing tech aids in reducing system size and space, while enhancing resolution and ensuring low latency for precise object detection.
