AI Edge Computing Module with Kneron KL520 NPU


  • Kneron KL520 NPU (Designed by ARM architecture)
  • M.2 B Key (22x80mm)
  • Accelerator for AI Edge Computing
  • Good for people monitoring, retail application

AAEON has partnered with Kneron to deliver modules featuring their innovative AI Edge Computing processor, the KL520 Neural Processor Unit (NPU). The KL520 provides efficient AI computing performance of 0.56 TOPS per Watt, perfect for remote, mobile, and unmanned applications.

The M2AI-2280-520 features the energy efficient KL520 in a M.2 22x80mm form factor that’s equally cost efficient. AAEON Kneron Modules are compatible with all of our embedded solutions which support M.2 22x80mm modules. The M2AI-2280-520 with Kneron KL520 NPU offers the best balance of performance, efficiency and cost.



IC Kneron KL520
Type Integrated SoC
Support Framework ONNX, TensorFlow, Keras, Caffe
Support Model Vgg16, Resnet, GoogleNet, YOLO, Tiny YOLO, Lenet, MobileNet, DenseNet
Memory Type LPDDR2
NPU Performance 0.35 TOPS
Overall Power Consumption 0.5 W


Operating Temperature 32°F ~ 158°F (0°C ~ 70°C)
Storage Temperature -40°F ~ 185°F (-40°C ~ 85°C)
Operating Humidity 0% ~ 90% relative humidity, non-condensing
Certification CE/FCC Class A
For packing list, detailed specifications and other product information, download the Datasheet or User Manual.