Open search

Edge AI Computing

Host controllers with CAN connectivity

Vecow (Taiwan) offers host controllers with isolated Classical CAN and CAN FD interfaces. The EAC-3000 is intended for in-vehicle applications, robot control, and automated guided vehicles.

The EAC-3000 controller is powered the Nvidia Carmel Arm central processing unit (Source: Vecow)

The Taiwanese company designs, develops, and produces high-end controllers for edge AI (artificial intelligence) computing. Recently, the company introduced the EAC-3000 controller. In Germany, the company is represented by Plug-In Electronic. Some of the supplied host controllers comes with up to two isolated Classical CAN or CAN FD interfaces. The applications addressed include rolling stock, commercial vehicles, robotics, autonomous driving among others. The products are fanless.

The EAC-3000 is built on the NVIDIA Jetson AGX Xavier platform featuring an 8-core NVIDIA Carmel ARM CPU, a 512-core NVIDIA Volta™ GPU, and up to 64 GiB LPDDR4x memory. Besides two CAN interfaces, the controller features four USB ports, two serial links, and five 1- Gbit interfaces. The device is also equipped with six antennas to support industrial-grade connectivity by M.2 Key E and Key B, allowing for 5G/4G/LTE/WiFi/BT/GPS communications. The product can be mounted on the wall mount. With anti-shock, anti-vibration and IEC 61373 certified, the host controller is suitable for harsh environment applications.

The company has also introduced the EAC-500 edge AI controller. It is equipped with the Nvidia Jetson AGX Orin system on module. It addresses similar applications as the above-mentioned product, but provides two isolated CAN FD ports. "The EAC-5000 series raises the bar for embedded computing applications with advanced AI performance, low-power consumption in a small form factor design,” said Esther Han, Product Manager at Vecow. “By leveraging the power of the latest NVIDIA Jetson AGX Orin, the Vecow EAC-5000 series is well-suited for server-grade IoT applications to address needs for intensive-computing, flexible expansion and trusted reliability.” The product will be sampled the in August this year. The product is scheduled for mass production in October.


Publish date

Plug-In Electronic