Edge AI PC GEAC91VP series

embeddedNVIDIA® Jetson Xavier™ NXCAN bus

The GEAC91VP A1 controller is designed specifically for the application of self-driving vehicles such as mobile robot, robotaxi, robobus, etc. With two NVIDIA Jetson Xavier modules inside, the controller is able to meet the critical computing power requirement from the complicated operations of multi-sersor data fusion,object classification,behavior forecast,control strategy determination, and so on. An automotive-grade multi-core MCIJ TC397 from Infineon,is embedded on the carrying board to execute the critical tasks of controlling the vehicle and managing the system status. A GPS module and a CPLD chip consist of the time synchronizing system, which provides time information and trigger signals aligning to the UTC time to different processors and sensors connected to the controller. The GEAC91VP controller supports most commonly used sensors on the self-driving vehicles, including GMSL cameras, LiDARs,radars, ultrasonic radars, etc. There are two versions of water cooling and air cooling available for the product. Product Characteristics Two NVIDIA Jetson Xavier modules vehicle-level MCIJ processor to run a safety system that meets the requirements of vehicle regulations The built-in hardware timing synchronization circuit provides a safe and reliable clock source for the autonomous driving vehicle system Main interface resources: provide 16 x GMSL2 camera interfaces, each can be triggered independently; 12 x CAN bus interfaces (all support canfd); 2 x LIN interfaces, 2 x Gigabit car Ethernet, 2 x 10 Gigabit Ethernet, supports for dual power supply, PPS, GPRMC, TRIGGER, P TP and other commonly used external timing interfaces