Edge AI Appliance

An edge AI appliance powered by AMD Kria™ K26 SOM and 4 Kinara Ara-1 processors that can run vision AI applications on 8 concurrent streams at high performance and a fraction of the cost compared to GPU-based solutions. Ideally suited for smart city and smart retail markets.

Video Provided by Kinara


  • Edge appliance supporting 8 concurrent video channels 
  • Support for leading-edge AI models (e.g., Efficientnet-D7, VIT)
  • Hardware-accelerated decoding, pre-, post-processing, and inference for 8 video channels
  • Full visibility with an SDK 
  • Automatic model optimization without user intervention


  • Retail: Frictionless checkout, shelf analytics, shrinkage reduction, virtual trial rooms
  • Smart security: People tracking, safety monitoring, crowd behavior detection
  • Smart transportation: License plate recognition, traffic monitoring, parking spot visualization
  • Smart manufacturing: Vision-based quality, warehouse robots, asset tracking

Kit Components:

  • Edge AI Box with thermal solution (heat sink and fan) 
  • 12V @1 amp power adapter with joiner

odm Partner Information


VVDN has more than a decade of experience in product engineering and manufacturing electronic solutions. VVDN helps customers steer their next-generation solutions with end-to-end design, development, testing, certification, and manufacturing – all under one roof. Their focus is to help customers with their next technology innovation in the space of 5G, Wireless, AI/ML on vision-based solutions, IoT, Cloud & Apps. To enable that, VVDN has set up R&D centers, testing lab infrastructure, and world-class manufacturing facilities in India.


Kinara is a leader in edge AI acceleration, building a cost and energy efficient edge AI inference platform supported by comprehensive software tools. Designed to enable smart applications across retail, medical, industry 4.0, automotive, and smart cities market segments, Kinara’s AI processors, modules, and software can be found at the heart of the AI industry’s most exciting and influential innovations. Led by Silicon Valley veterans and a world class development team in India, Kinara envisions a world of exceptional customer experiences, better manufacturing efficiency, and greater safety for all.

Frequently Asked Questions

Yes, VVDN provides BSP-level software support along with Kinara and the PetaLinux SDK in the box. We also offer customized design services to cater to the custom design requirements of the customer on a need basis.

Tutorials for model conversion and bring-up are available at the Kinara.ai SDK portal (requires NDA).

PyTorch, TensorFlow, Caffe, MXNet, and ONNX.

There is support for all state-of-the-art models, including 2D CNN, 3D CNN, vision transformers, and so on. The full list of models is available from Kinara under NDA.

Box dimensions are: 205mm (w) X 255.2mm (d) X 42mm (h). Please contact VVDN if you require a different form factor.

Expansion and customization options are available under NRE. 

Featured Documents