Subscribe to Our Newsletter

Success! Now Check Your Email

To complete Subscribe, click the confirmation link in your inbox. If it doesn't arrive within 3 minutes, check your spam folder.

Ok, Thanks
Raspberry Pi got a new HAT+ adapter with an integrated Hailo accelerator
Credit: Raspberry Pi

Raspberry Pi got a new HAT+ adapter with an integrated Hailo accelerator

Raspberry Pi has launched the AI HAT+, a new AI accelerator board available in 13 TOPS ($70) and 26 TOPS ($110) variants, featuring integrated Hailo technology that enables complex neural network processing for Raspberry Pi computers.

Ellie Ramirez-Camara profile image
by Ellie Ramirez-Camara

Raspberry Pi recently expanded its AI product line with the Raspberry Pi AI HAT+ launch. The new adapter comes in two flavors: one featuring an integrated Hailo-8L AI accelerator—the same as in the AI Kit— reaching up to 13 TOPS and with a $70 price, and a version including a more powerful 26 TOPS Hailo-8 accelerator, costing $110. Unlike the AI Kit, the HAT+ does not require an M.2 connector, as the chip is integrated directly into the main PCB for a simpler design that dissipates heat better.

The AI HAT+ supports PCIe Gen 3.0 mode to optimize the Hailo-8 accelerator's performance. Hailo-8's 26 TOPS enable AI workloads not accessible to the Hailo-8L setups, such as the possibility of running more sophisticated neural networks, or simultaneously executing multiple neural networks at high frame rates to allow performing complex tasks like object detection, pose estimation, and subject segmentation on a live camera feed. Both versions maintain backward compatibility with existing Raspberry Pi AI Kit software, ensuring a smooth transition for developers already working with Hailo accelerators.

Users of the new AI HAT+ should find no compatibility issues, since this product is backward compatible with the AI Kit, and needs no modifications to integrate with Raspberry Pi's camera software stack. The only outstanding issue could be that models designed for the Hailo-8 may not run on the Hailo-8L, but alternatives are likely available. In any case, Raspberry Pi's continued improvement of its AI-focused offerings makes it a remarkable hardware host for multiple types of AI workloads.

Ellie Ramirez-Camara profile image
by Ellie Ramirez-Camara
Updated

Data Phoenix Digest

Subscribe to the weekly digest with a summary of the top research papers, articles, news, and our community events, to keep track of trends and grow in the Data & AI world!

Success! Now Check Your Email

To complete Subscribe, click the confirmation link in your inbox. If it doesn’t arrive within 3 minutes, check your spam folder.

Ok, Thanks

Read More