NVIDIA Jetson TX2 AI computing platform for AIoT

The NVIDIA Jetson TX2 pairs a dual-core Denver 2 alongside a quad-core ARM Cortex-A57 processor and provides 256 CUDA cores on the NVIDIA’s advanced Pascalä GPU architecture with up to 1.33 TFLOPS, delivering exceptional AI performance. 

Designed to operate under an extended temperature range from -20 to +60°C, the EC-3200 leverages the integrated GPU-computing power of the Jetson TX2 platform in building edge inference servers to analyze and better manage real time traffic flow in smart cities or to optimize the efficiency of operational processes in intelligent factories.

“Current edge and cloud-based AI products require better computational and video analytics capability to perform demanding real-time data processing and overcome latency issues,” said Wilson Lin, Director of IBASE Product Planning Department. “The EC-3200 is no doubt a durable solution, especially built with a fanless design for non-stop operation and takes advantage of the GPU-accelerated parallel processing of NVIDIA Jetson TX2 to handle data intensive and mission-critical workloads with power efficiency and unmatched reliability.”

With dimensions of 127x137x46mm, the EC-3200 comes with 8GB LPDDR4 memory and 32GB eMMC storage on board. To connect to essential peripherals, front I/O interfaces provide a GbE, an HDMI (1.4), two USB 3.0 and two USB 2.0 ports, while a DB9 connector (For RS232 or CAN bus), a Micro USB slot, as well as an external 10-pin GPIO port are located on the right side.

The two M.2 expansion slots are designed to support optional SSD storage (M2280) and WAN module (E3042). The system is outfitted with a total of four external antenna ports to enhance throughput and connection reliability in outdoor environments. The OS is based on a custom made Ubuntu 16.04 with Jetpack 3.2.1 and L4T 28.2.