EdgeKit - A modular Edge AI Computer based on NVIDIA Jetson

EdgeKit is a highly modular edge AI computer powered by NVIDIA® Jetson™ product family. It powers AI software solutions with a hardware platform that perfectly fits every use-case without long engineering lead times and high costs.

  • Supports NVIDIA Jetson product family (Xavier NX, Orin Nano, and Orin NX)
  • Compatible with FPD, Ethernet, or CSI cameras for real-time machine vision tasks
  • Select more than 40 different options for the best use-case fit using an interactive web configurator

 

Interfaces

Interface Optional Specifications
NVIDIA® Jetson connector no 260-pin SO-DIMM, supported SOMs are listed here
Micro USB no USB 2.0, 1 micro type-B slot, OTG not supported
Micro SD card slot no Not compatible with Orin SOMs
USB no

USB 3.0 (Nano and TX2 NX), USB 3.1 (Xavier NX and Orin NX), shared, 2 or 4 slots (optional). 

Important notice: USB Cameras are not recommended, since there are a lot of driver incompatibilities. Use Ethernet of FPD Link instead.

Ethernet no RJ45 connector, 1 Gbps, 1 or 2 slots (optional)
HDMI no HDMI 2.0, supports resolutions up to 3840x2160 (4K)
CSI no 2-lanes, 2 or 4 slots (optional), Xavier NX supports up to 6 slots (optional)
GPIO no 40-pin header including 1x UART, 2x I2C, 1x I2S and 2x SPI
Debug header no UART on baseboard, 3.3 V, Force Reset pin
SATA yes Not compatible with 4x USB 3.1
SSD yes M.2 M-Key socket, optional built-in SSD with 128 - 1024 GB storage size
WiFi yes M.2 E-Key socket, built-in module, antenna included
GPS yes M.2 B-Key socket, built-in module (U-blox NEO M8 series GPS module), antenna included
LTE yes mini PCI-Express socket, built-in module, antenna included
CAN yes M12 connector, isolated, software switchable termination
Relay yes 2x 3-pin, plugable screw terminal, SPDT output, max. 50 VDC, max. 1 A
Opto-isolated IO yes 4x optoMOS output, max. 24 V
4x input, max. 24 V
Serial yes D-SUB connector, software selectable RS232/422/485, software switchable termination, max. 20 Mbps

Supported NVIDIA® Jetson Boards

Jetson Xavier NX Jetson Orin Nano Jetson Orin NX
AI Performance 21 TOPS 20/40 TOPS 70/100 TOPS
GPU 384-core NVIDIA Volta™ GPU with 48 Tensor Cores 1024-core NVIDIA Ampere GPU with 32 Tensor Cores 1024-core NVIDIA Ampere GPU with 32 Tensor Cores
CPU 6-core NVIDIA Carmel ARM® v8.2 64-bit CPU 6-core ARM® Cortex® A78AE v8.2 64-bit CPU 6-core / 8-core Arm® Cortex® A78AE v8.2 64-bit CPU 1,5 MB / 2MB L2 + 4 MB L3
RAM 8GB 128-bit LPDDR4x 16GB 128-bit LPDDR4x 4 GB 64-bit LPDDR5 / 8 GB 128-bit LPDDR5 8 GB 128-bit LPDDR5 / 16 GB 128-bit LPDDR5
Storage 16GB eMMC 5.1 Supports external NVMe
Power consumption 10 - 20W 5W - 10W | 7W - 15W 10W - 20W | 10W - 25W

Technical specifications

Voltage 12, 24, 48 VDC
Power consumption max. 60 W
Ambient temperature 0 °C to 50 °C
Humidity Not condensing
Cooling max. 15 W passive cooling
optional air cooling with 2x 60 mm fans
Weight 1 kg
Housing Black eloxated aluminum
Installation optional, DIN-Rail or Wall mount


Legal notice

European Union

The low band 5.15 -5.35 GHz is for indoor use only.

This equipment complies with the essential requirements of the European Union directive 2014/53/EU. See Statements of European Union Compliance. European Union Declarations of Conformity.

IMPORTANT: EdgeKit is for commercial use only. 

Mechanical specifications

Low-profile variant (without extension boards)

Width: 136.0 mm
Height: 52.5 mm
Depth: 83.0 mm

Extended variant (with industrial or media extension board)

Width: 136.0 mm
Height: 73.5 mm
Depth: 83.0 mm

Scope of delivery

  • EdgeKit with latest version of NVIDIA® JetPack SDK pre-installed
  • Print version of quick guide and CE declaration of conformity
  • Antennas for GPS, LTE or WiFi interfaces (if present)
  • Connector for 2-pin terminal (if present)
  • Connector for opto-isolated IO (if present)
  • Connector for relay (if present)

Downloads

PDF: CE declaration of conformity (radio equipment directive) PDF: CE declaration of conformity (EMC directive) PDF: CE declaration of conformity (simplified) + Quick Guide PDF: User manual Link: GitHub repository