RASPBERRY PI AI HAT+ 26 TOPS, RASPBERRY PI SC1791

Description

The Raspberry Pi AI Kit is an innovative solution designed to enhance the capabilities of the Raspberry Pi 5 by integrating advanced artificial intelligence (AI) functionality. This kit includes the Raspberry Pi M.2 HAT+ paired with a Hailo AI acceleration module, creating a powerful and efficient AI platform. The Hailo module, featuring a 26 TOPS neural network inference accelerator with the Hailo-8L chip, utilizes the M.2 2242 form factor and is pre-installed on the M.2 HAT+. This setup enables a seamless connection through an M key edge connector, linking the AI module's M.2 interface with the Raspberry Pi 5’s PCIe 2.0 interface.

Upon running an updated Raspberry Pi OS image, the Raspberry Pi 5 automatically recognizes the Hailo module, activating the Neural Processing Unit (NPU) for AI computing tasks. Additionally, the Raspberry Pi OS's rpicam-apps camera applications are configured to leverage the NPU for running AI-driven post-processing tasks, enhancing the system’s overall functionality without requiring manual configuration. This integration offers an accessible, cost-effective way to explore high-performance AI with Raspberry Pi 5.

Specifications:

• Performance: Contains a neural network inference accelerator capable of delivering 26 tera-operations per second (TOPS), enabling high-speed AI computations.
• Integration: Fully integrated into the Raspberry Pi’s camera software stack, allowing for seamless operation with camera-based applications.
• Thermal Management: Includes a pre-fitted thermal pad between the module and the HAT+, which helps distribute heat evenly across components to enhance performance and reliability.
• Compatibility: Conforms to the Raspberry Pi HAT+ specification, ensuring a perfect fit and optimal functionality with the Raspberry Pi ecosystem.
• Hardware Inclusions: Comes with a 16mm stacking header, spacers, and screws, facilitating easy installation on the Raspberry Pi 5, even with the Raspberry Pi Active Cooler attached.

Links: Product_Brief (PDF)