Home / Products / IoT Communication Devices / Edge Computing Gateway / Edge AI Gateways /

GB400 Industrial Gateway



◆ NVIDIA Jetson AGX Xavier 

◆ Efficient fanless heat dissipation without frequency reduction

◆ IP50 protection

◆ Flexible expansion, multi-sensor access

◆ Video storage, AI analysis

◆ 10G high seismic resistance, stable system

◆ Deep learning acceleration

◆ Easy to maintain

◆ Quick upgrade


Collect it   + Follow updates Get a quote

Company name

Contact person(Optional)



Contact us Overview Specifications Resources Applications

The GB400 industrial gateway is equipped with NVIDIA Jetson AGX Xavier core module, containing 512 CUDA cores, Volta architecture GPUand NVDLA engine. It boasts 32 Tops real-time reasoning capability with 30W module power consumption so as to be compatible with all current neural network models. The whole machine features fanless and strong embedded design. It has strong adaptability to harsh industrial environments thanks to the wide temperature working environment and high seismic protection level. 

The GB400 industrial gateway integrates a variety of expansion interfaces, and supports 4G/5G/Wi-Fi communication. The standard PCIex8reserved design fully interprets its high scalability that allows the access of a variety of special sensors. It is widely used for smart city vehicleroad collaboration, assisted driving, unmanned delivery vehicles, intelligent inspections and other scenarios.

Its application in vehicle-road synergy and assisted driving enables massive terminal access management to improve operation and maintenance efficiency, reduces the difficulty of integration with the open and compatible platform, and facilitates the processing of computing tasks. For the auxiliary computing platform in high-speed autonomous vehicles, customers can carry out secondary development. By processing image data, combined with the capabilities of cameras and radars, it can complete the perception and detection of road events such as lane line recognition, traffic light recognition, pedestrian detection, and driver fatigue detection and analysis.

Processor NVIDIA Jetson AGX Xavier
AI performance 32TOPS/11TFLOPS
CPU 8-Core Carmel ARM V8.2 64-Bit CPU, 8MB L2 + 4MB L3
GPU 512-Core NVIDIA Volta GPU with 64 Tensor Cores
Memory 32GB 256-Bit LPDDR4x
Storage 32GB eMMC 5.1
Ethernet 2 x Gigabit Ethernet
CAN 2 x CAN (DB9)
Video encode 4 x 4K @ 60 (HEVC) 16 x 1080p @ 60 (HEVC) 32 x 1080p @ 30 (HEVC)
USB 2 x USB3.1 1 x USB Type C
Display 1 x HDMI
Audio Headset Mic-in, headphone out
Button 1 x Power button, 1 x Recovery button
Serial 2 x RS232/485 (DB9)
User expansion ① 1 x Mini-PCIe, ②1 x M.2 E-Key 2230 for Wi-Fi (Alternatively optional) 1 x M.2 M-key 2280 for NVME (PCIE) 1 x M.2 B-KEY for LTE/5G 1 x TF slot 2 X SATA 1 x Nano SIM Socket 2 x GMSL
DLA accelerator 2 x NVDLA Engine
Dimensions 256 x 174 x 87 mm
IP rating IP50
Net weight 1.2 kg
Input 12-36V DC
Temperature Operating: -20°C ~ 60°C CPU/GPU Full Loaded Operation - 20°C ~ 75°C Frequency Reduction Operation Storage: -40°C ~ 70°C
Humidity RH 5%~95% @ 40°C (Non-condensing)
Operation vibration 3 Grms 5~500Hz

Supported documents
Name Type Version Release date

GB400 Industrial Gateway Datasheet

Document 2021/11/11

Subscribe Vantron Newsletter