Raspberry Pi, the renowned maker of affordable computing solutions, has recently ramped up its product offerings, unveiling a series of innovative add-ons designed to enhance the capabilities of its flagship Raspberry Pi 5. Known for their versatile and compact single-board computers utilized by hobbyists, educators, and industry professionals alike, the Raspberry Pi products continue to evolve, catering to the growing interest in artificial intelligence and data processing.
The Raspberry Pi 5 features a PCIe 3.0 interface, which is accessed through a 16-pin connector, allowing users to expand their computing experience. Building on this foundation, Raspberry Pi has launched M.2 HAT+ extension cards that convert the 16-pin interface into a more commonly used M.2 format. These add-ons, known as HATs (Hardware Attached on Top), allow users to attach various modules and accessories, making the Raspberry Pi 5 even more adaptable to different applications.
One prominent example of how users are leveraging the M.2 slot is by adding NVMe SSDs to enhance data storage and retrieval speeds. In June, Raspberry Pi introduced an AI Kit that utilized an M.2 extension card integrated with a neural network inference accelerator from Hailo, showcasing the potential of these tiny computers in AI applications. Now, the company takes this initiative a step further by announcing a new AI HAT+ addition, featuring a built-in Hailo inference accelerator specifically designed for use with Raspberry Pi 5.
The new Raspberry Pi AI HAT+ comes in two variants, offering either 13 or 26 tera-operations per second (TOPS) capabilities, priced at $70 and $110 respectively. The entry-level option boasts the same powerful module as the one previously included in the AI Kit, emphasizing affordability without compromising performance. While it’s essential to note that users will not be training AI models like GPT on these devices, these add-ons present a cost-effective solution for running inference tasks at the edge, making AI more accessible to the broader community.
Unlike the earlier M.2 HAT+, the AI HAT+ does not feature its own M.2 interface, being designed strictly for Hailo’s inference modules. This streamlined approach simplifies the integration process for users who are eager to explore AI capabilities with their Raspberry Pi 5 systems.
Additionally, Raspberry Pi is making strides in the storage domain by entering the SSD market with its own branded M.2 NVMe SSDs. The new offerings include a 256GB variant priced at $30 and a 512GB version for $45. These prices compete directly with off-the-shelf alternatives available on platforms like Amazon, which currently list similar SSD specifications between $20 and $30 for the 256GB model. For those looking to simplify their purchase, Raspberry Pi is also offering SSD Kit bundles that include both the M.2 HAT+ and a compatible SSD, available for $40 for the 256GB kit and $55 for the 512GB counterpart.
While these SSD Kits may not spark a significant revolution within the Raspberry Pi ecosystem, they serve a practical purpose, ensuring compatibility with the Raspberry Pi 5 and its M.2 HAT+ add-ons. The company’s announcements highlight a discerning step forward into the realm of artificial intelligence and efficient data management, making Raspberry Pi a formidable player in the tech landscape.