XaLogic had launched a campaign on Crowdsupply for a Kendryte K210 based HAT. The K210 AI Accelerator is a compact Raspberry Pi HAT that utilizes the Kendryte K210 AI processor to enable 0.5 TOPs (Tera Operations Per Second) of processing power. Kendryte K210 is a dual-core RISC-V AI processor that was launched in 2018, and you can find it in several smart audio and computer vision solutions. The HAT also supplies a Trust-M security chip. The 65 x 30mm HAT is designed to be compatible with any 40-pin Raspberry Pi equipped with an RPi camera add-on. The HAT enables you to add AI features to your RPi based camera even if you don’t know how to train your model. The plugin module, as well as the pre-trained models, will make your camera AI-enabled in minutes with a few Python API calls. Its Kendryte K210 SoC provides 0.5-TOPS NPU performance and can run at as low as 0.3W. This is in contrast to Coral Edge TPU which clocks at 2W, and 1.5W for Intel’s Myriad X powered Neural Compute Stick 2, with both of them running at about 4 TOPS.

XaLogic K210 AI Accelerator functions best with Raspberry Pi Zero and camera and enables you to make use of pre-trained models for evaluation such as object detection, face detection, age, and gender estimation,  simple voice commands, and vibration abnormally detection. You can also train your own model using a more powerful host machine. You can make use of an NVIDIA GPU, with TensorFlow. You can make use of TFLite, Caffe, and even ONNX format through the aid of conversion tools. You can make use of Visual Studio Code, which is recommended by XaLogic to modify K210 C code directly on Raspberry Pi if need be. With the aid of the Visual Studio Code for Raspberry Pi and the pivotal toolchain for the K210, you can develop all the K210 firmware on the Pi itself. The K210 AI Accelerator features an Infineon Trust-M onboard which enables you to establish a secure connection to AWS through MQTT without exposing the private key. This feature is important for your privacy when you are deploying your IoT devices in the field.

Schematics, C code in the K210, and all code running on the Raspberry Pi will be open-sourced. Pre-trained models are provided in binary form. Also, sample Caffe and Tensorflow projects are available to help you create your own custom neural network.


About The Author

Muhammad Bilal

I am highly skilled and motivated individual with a Master's degree in Computer Science. I have extensive experience in technical writing and a deep understanding of SEO practices.

Scroll to Top