Edge AI Inference System with Intel Atom® E3940 QC SoC and Two Intel MA2485 VPU

Anewtech AD-AIR-101 AI Inference System Advantech Embedded PC


  • Intel Atom E3940 Quad Core SoC turbo burst up to 1.8 GHz
  • Intel® Movidius™ Myriad™ X VPU MA2485 x 2 Built-in
  • Intel® OpenVINO™ toolkit supported
  • SATA Slim 64G SSD x1 Built-in
  • 204-pin SO-DIMM DDR3L 8GB memory x 1 built-in
  • DIN-Rail AI system with essential I/O ports on front-side bezel
  • M.2 2230 for WIFI or optional VEGA-320 module
  • 12V ~ 28V wide-range power input
  • Integrated with Edge AI Suite



Anewtech air-101 AI Inference System Advantech Embedded PC

Application: AI-driven Fruit Grading and Inspection

Traditional fruit grading and inspection — with human eyesight or automated grading machines — is time-consuming, labour-intensive, and costly. Fortunately, innovative AI technologies improve inspection efficiency and precision. Fruit sent to grading machines are scanned by IP cameras and sorted according to appearance, size, and colour according to an AI model. This fruit is then inspected using other methods to ensure quality.

This customer leveraged an AD-AIR-101 AI inference system and VPU vision processor for fruit grading. This reliable system provided efficient visual computing performance, multiple I/O interfaces, and DIN rail design to deliver high precision grading and allow customers to easily deploy.



  • Low power consumption and high scalable VPU, more affordable than GPU and FPGA solutions
  • Edge AI Suite and Intel® OpenVINO™ toolkit accelerate deployment
  • Rich I/O interface for various connected devices — including cameras, scale sensors, code scanners, and printers
Model AIR-101-S62A1
CPU Intel Atom E3940 Quad Core Soc
Frequency 1.6 GHz, turbo burst 1.8GHz
Technology DDR3L 1866 MHz
Default (Built in) 1 x 204-pin SO-DIMM DDR3L 8GB memory (-40 ~ 85 °C) built-in
AI Engine
VPU Two Intel® Movidius™ Myriad™ X MA2485 (1 x VEGA-330-02A1 Built-in)
Software Advantech Edge AI Suite, OpenVINO R2 2019 or latest
Chipset Intel HD Graphics 500, CPU integrated
HDMI HDMI1 : 3840 x 2160 @ 30Hz; HDMI2 : 3840 x 2160 @ 30Hz
LAN1 10/100/1000 Mbps Intel i210 GbE, support Wake On Lan
LAN2 10/100/1000 Mbps Intel I210 GbE, support Wake on LAN
Interface Realtek ALC888S, High Definition Audio, Line-out, Mic-in
I/O Interface
Serial Port 2 x RS-232/422/485 with auto flow control, COM2 with 5V/12V power supply
USB 4 ports (USB3.0)
GPIO 8-bit programmable DIO (optional)
M.2 1 x M.2 2230 E key for WIFI or VEGA-320 module
SSD 1 x SATA Slim 64G SSD (-40 ~ 85 °C) built-in
Operating System
Microsoft Windows Windows 10 Enterprise 64 bit
Linux Ubuntu 16.04 LTS (64 bit)
Power Requirement
Input Voltage 12 ~ 28 VDC
Power Adapter AC to DC 24V/2.5A 60W (Optional)
Power Consumption
Typical 8.7W
Max. 19.2W
Construction Aluminum housing
Mounting DIN-rail/Wall Mount
Dimensions 53.5 x 158 x 114 mm
Operating Temperature With built-in VEGA-330: -20 ~ 55 °C, with 0.7m/s air flow; With built-in VEGA-330 and optional VEGA-320: -20 ~ 50 °C, with 0.7m/s air flow
Safety Certifications CB, UL, CCC, BSMI

Edge AI Suite Features: