Electrification, infotainment and advanced driver-assistance systems (ADAS) are changing our conception of what cars can be. These innovations, however, are creating many new design challenges that require sophisticated technology to overcome. Increasingly, these designs need parallel processing and real-time performance to support the many new onboard sensors and complex multi-camera systems.
Xilinx is responding to these requirements with automotive-qualified devices that provide the programmable capacity, power efficiency and high I/O to support these systems. These new devices feature a range of differentiators including AI integration, system level reliability and determinism, a broad set of any-to-any interfacing capabilities, and scalar and adaptable processing engines for the most demanding ADAS applications.
Addressing these needs, Xilinx recently introduced the automotive qualified Zynq UltraScale+ MPSoC family, enabling development of safety critical ADAS and autonomous driving systems. The family is designed to meet AEC-Q100 test specifications and full ISO 26262 ASIL-C level certification.
In addition, the new Virtex-4 platform FPGAs are the industry’s lowest-cost automotive FPGAs and feature a wide range of third-party IP that provides CAN support for gateway, infotainment, and driver assistance applications. This new IP is available in the Xilinx LogiCore library and optimized for XA devices. This IP offers flexible CAN protocol implementation with up to 64 message buffers on both the transmit and receive sides of the chip. This provides designers the flexibility to choose between a wide variety of protocol implementations while minimizing development time and risk. Xilinx also is collaborating with several partners to accelerate the deployment of full-stack ADAS software and hardware systems for parking & cruising, front view & surround view and autonomous control functions.