Note: It is highly recommended to get a proper cooling mechanism(Active or Passive) and a proper PD (Power Delivery 5A@5V) PSU for the Raspberry Pi 5 and the AI kit.
The 1st Official AI Kit for Raspberry Pi 5
The wait is over! Raspberry Pi officially enters the AI revolution with the Raspberry Pi AI Kit - 13 TOPS (Trillion Operation Per Second or Tera Operation Per Second) neural network inference accelerator for Raspberry Pi 5. This official product empowers you to unlock the incredible potential of AI for your projects, especially with Raspberry Pi 5.
Check out our live demo:
Unparalleled Performance with HAILO 8L AI Accelerator
Dive headfirst into the world of AI with the Raspberry Pi AI Kit. At its core lies the mighty Hailo 8L AI accelerator, boasting a staggering 13 TOPS (Trillion Operations Per Second) inferencing capability. This translates to lightning-fast inference speeds, ideal for handling complex AI tasks in real-time at the edge. Whether you're building an intelligent image recognition system, experimenting with object detection, or mapping with a mobile robot, the Hailo 8L ensures your projects run smoothly without sacrificing performance. But the Hailo 8L goes beyond raw power. It claims to be the most efficient edge AI processor in the market, achieving an order of magnitude better performance in both area and power consumption than competitors. This efficiency is crucial for running deep learning models on resource-constrained, battery-powered, edge devices like Raspberry Pi 5. Additionally, the Hailo 8L utilizes a specialized architecture optimized for neural networks which is the Neural Processing Unit (NPU), housed in a compact M.2 form factor for seamless integration. It operates within a wide temperature range (-40ºC to 85ºC) and offers scalability to handle multiple data streams and models simultaneously, making it a versatile solution for various AI applications.
Check out how Jeff Geerling uses this AI kit with Raspberry Pi 5, and Camera Module and demonstrate some amazing computer vision:
Another good overview of this AI kit and assembling it onto the Raspberry Pi 5, and demo at the end, by Kevin McAleer:
Here is the comparison that we (Cytron Technologies) list out, comparing the Hailo 8L - 13 TOPS AI Accelerator with other popular AI accelerators:
Seamless Integration and Hassle-Free Development
Gone are the days of complex setup procedures. The Raspberry Pi AI Kit prioritizes user-friendliness. It comes pre-assembled on a Raspberry Pi M.2 HAT+, eliminating the need for intricate assembly processes. This HAT+ design ensures flawless compatibility with your Raspberry Pi 5, allowing you to plug it in and start developing right away. Furthermore, the kit integrates seamlessly with Raspberry Pi OS and the existing camera software stack (libcamera, rpicam-apps, and picamera2). This means you can leverage existing libraries and tools within your familiar development environment, saving valuable time and streamlining the creation process. And of course, it is also quite straightforward to install the driver and software package (HailoRT, Hailo Tappas, and more) for the Hailo-8L from the existing Raspberry Pi OS using the apt package manager. And before you ask, No, the HAILO 8L AI Accelerator is NOT SOLD separately, it comes pre-assembled with the Raspberry Pi M.2 HAT+.
Check out some of the demos from Raspberry Pi: (All are using Raspberry Pi 5 with the AI kit)
Object recognition: even works when traffic on the A14 is moving
Pose estimation: very casual, not at all suspicious
Object recognition: the lemon Coke “bottle” is controversial in these parts
You can also checkout the video by Hailo team, demostrating installation of Hailo Driver, HailoRT, Hailo TAPPAS, and also applications:
The HAILO 8L AI Accelerator offers:
- Performance: Up to 13 TOPS (Tera Operations Per Second) processing capability, making it efficient for real-time AI tasks.
- Efficiency: Hailo-8L claims to outperform all other edge AI processors in area and power efficiency by an order of magnitude.
- Architecture: Utilizes a structure-driven dataflow architecture to optimize neural networks.
- Form Factor: M.2 form factor for easy integration with existing devices that have an M.2 socket with B or M-key.
- Power Consumption: Low power consumption allows for running deep neural network inference on edge devices.
- Temperature Range: Supports a wide temperature range of -40°C to 85°C for various environments.
- Scalability: Enables simultaneous processing of multiple data streams and models, making it suitable for complex AI applications.
Perfect for Computer Vision and Edge AI Robotic
BTW, the AI kit is fully compatible with the latest Raspberry Pi OS (Bookworm) and also all the camera modules and the software stack. We are very excited about the 1st AI Kit from Raspberry Pi, It empowers many of us to:
- Develop cutting-edge AI applications:
- Image recognition and classification
- Object detection and tracking
- Facial recognition
- Natural Language Processing (NLP) tasks (with additional software)
- And much more!
- Elevate your projects:
- Add intelligent features to your Raspberry Pi creations
- Experiment with real-world AI applications
- Push the boundaries of creativity and innovation
You will see a leap improvement in existing computer vision projects based on Raspberry Pi, for example, the Donkeycar, and OpenCV image processing projects.
The Raspberry Pi AI Kit - 13 TOPS for Raspberry Pi 5 is the perfect solution for developers, hobbyists, and educators seeking to explore the exciting world of AI. Join the revolution and take your Raspberry Pi projects to the next level!
What Will You Need to Get Started?
As the AI kit is designed for Raspberry Pi 5, you will of course need to get a Raspberry Pi 5 and accessories. Here are our recommendations for you.
Note: We highly recommended to get a proper cooling mechanism(Active or Passive) and proper PD (Power Delivery) PSU for the Raspberry Pi 5.
- The mainboard itself:
- A good 27W PD (Power Delivery) Power Supply or Power Adapter - Because Raspberry Pi 5 is more powerful and to get the most out of Raspberry Pi 5 and the AI kit, please get yourself a true 27W PD PSU that can deliver 5A @ 5V.
- Raspberry Pi 27W PD UK Plug White Color, suitable for Malaysia, Singapore, Brunei, Cambodia, Sri Lanka, UAE
- Raspberry Pi 27W PD UK Plug Black Color, suitable for Malaysia, Singapore, Brunei, Cambodia, Sri Lanka, UAE
- Raspberry Pi 27W PD EU Plug White Color, suitable for Thailand, Indonesia, Vietnam, and the Philippines
- Raspberry Pi 27W PD EU Plug Black Color, suitable for Thailand, Indonesia, Vietnam, and the Philippines
- Cytron 27W PD 5V 5A USB-C for Raspberry Pi - UK Plug, suitable for Malaysia, Singapore, Brunei, Cambodia, Sri Lanka, UAE
- Cytron 27W PD 5V 5A USB-C for Raspberry Pi - EU Plug, suitable for Thailand, Indonesia, Vietnam, Philippines
- Cytron 27W PD 5V 5A USB-C for Raspberry Pi - US Plug, suitable for US, Canada, Mexico, Japan, Thailand
- As the AI kit will utilize the PICe bus, the Raspberry Pi 5 can only boot from a microSD card or a USB drive. We recommend using the MakerDisk microSD card preloaded with Raspberry Pi OS - Bookworm. It will boot up out of the box:
- 256GB MakerDisk A2 Class 10 microSD - Pre-installed Raspberry Pi OS, in case you need large storage to store data such as videos, photos/pictures, and audio for AI Training.
- 128GB MakerDisk A1 Class 10 microSD - Pre-installed Raspberry Pi OS, in case you need large storage to store data such as videos, photos/pictures, and audio for AI Training.
- 64GB MakerDisk A1 Class 10 microSD - Pre-installed Raspberry Pi OS, 64GB suitable for AI model deployment.
- 32GB MakerDisk A1 Class 10 microSD - Pre-installed Raspberry Pi OS, 32GB suitable for AI model deployment.
- separately. Of course, if you have time and fast Internet, open up your laptop download the latest Raspberry Pi OS, and load it into your existing micro SD.
- Raspberry Pi 5 is more powerful and with higher power, it generates more heat during operation, especially with AI development and deployment. We recommend getting a good active heatsink which can integrate with the M.2 HAT+:
- With dual HDMI ports and limited space, the Raspberry Pi 5 can only utilize a micro HDMI socket. You can grab the:
- An enclosure is always good to protect the most important part of your computer. Since there are some changes in Raspberry Pi 5 compared to the previous boards, a new case is needed. Grab the official case:
FAQ for Raspberry Pi AI Kit
- Can I use the Raspberry Pi AI Kit with the Raspberry Pi 4 Model B?
Answer: No, because the AI kit required a PCIe socket for the interface/connection and so far, only Raspberry Pi 5 has that socket. - Can I use Raspberry Pi OS (Buster or Bullseye) older versions of Raspberry Pi OS work with Raspberry Pi AI Kit?
Answer: You will need the latest version of Raspberry Pi OS, Bookworm (64-bit) because the AI Kit is meant for Raspberry Pi 5, and Raspberry Pi 5 can only boot up with Bookworm. - Will my Raspberry Pi 4 power supply (3.0A or 3.5A @ 5V) work with the Raspberry Pi AI Kit?
Answer: The official Raspberry Pi 15W USB-C PSU or other compatible PSU for Raspberry Pi 4 can technically power the Raspberry Pi 5 and the AI kit, it might not be sufficient due to the Raspberry Pi 5's increased performance, and not to forget the AI kit requirement. For optimal performance and stability, we strongly recommend the official 27W high-quality 5V 5A USB-C power supply for Raspberry Pi 5 and this AI kit. - Can I get the Hailo 8L AI Accelerator separately?
Answer: No, the Hailo 8L AL Accelerator comes pre-assembled on M.2 HAT+. - Is the active cooler needed and fit with the AI kit?
Answer: Raspberry Pi 5 is faster and more powerful than prior-generation Raspberry Pis, and like most general-purpose computers, it will perform best with active cooling. The Raspberry Pi Heatsink & Active Cooler is a good cooling solution for the AI kit. - Can I boot up Raspberry Pi 5 from NVMe SSD while using the AI kit?
Answer: If you plan to use the PCIe bus for the NVMe SSD, then it will be NOT Possible because the AI kit needs the PCIe bus. Yet, you can still get a USB 3.0 to NVMe M.2 SSD adapter and boot the Raspberry Pi 5 from the USB 3.0 port. - Will the Raspberry Pi AI Kit fit the Official Raspberry Pi 5 case?
Answer: Yes, but you will need to remove the top lid and the fan.
International
Singapore
Malaysia
Thailand
Vietnam
.jpg)
