Raspberry Pi has introduced a new Raspberry Pi AI Kit, which makes advanced AI capabilities more accessible to hobbyists and developers. This kit combines two key components, on the one hand the official M.2 Key M HAT+ module, and on the other the Hailo-8L M.2 AI accelerator to be inserted into the HAT in the M.2 slot. This means that the Raspberry Pi can have up to 13 TOPS for AI processing, which is not bad at all to start with your first artificial intelligence projects executed locally.
The kit can be found for a price of about 70 dollars, a little less than 70 euros. Additionally, work has been done to improve availability, thanks to the involvement of Raspberry Pi, the kit should be more available than the individual components. On the other hand, you will find simplified documentation, to make it easier for you to get started in the field of AI.
While the hardware itself may not be innovative, the collaboration offers interesting software possibilities. A range of AI-powered computer vision samples are now available on the Raspberry Pi website and GitHub. These samples include object detection, pose estimation, and instance segmentation, all optimized for Raspberry Pi 5.
El The biggest benefit of using an AI accelerator like the Hailo-8L lies in its efficiency. Compared to the Raspberry Pi 5's CPU or GPU, the Hailo-8L consumes less power and offers significantly faster AI processing. An interesting "little toy" for enthusiasts and developers in the world of AI who already have a Raspberry Pi 5.
With the Raspberry Pi AI Kit, enthusiasts and developers now have a powerful and affordable tool to explore the world of AI on their Raspberry Pi 5. This opens the doors to a wide range of creative projects and innovative applications.
Raspberry Pi AI Kit technical specifications
As for the technical specifications of the new Raspberry Pi AI Kit, we have:
- SBC supported: Raspberry Pi 5
- M.2 HAT+ with PCIe Gen2 x1 interface, M.2 Key M slot
- M.2 module with Hailo-8L AI accelerator chip:
- Performance up to 13 TOPS
- M.2 2242 form factor
- 1.5W power consumption
- Thermal pad between the module and the HAT
- Hardware mounting kit included
- Stacked 16mm GPIO header
- PCIe FPC Cable