Accelerator Google Coral Edge TPU ASIC M.2 A+E key coprocessor to Frigate

The Coral M.2 Accelerator A+E key module is a device that enables the integration of the Edge TPU coprocessor from Google into existing systems, significantly improving the performance of machine learning (ML) inference processes while maintaining low power consumption. This coprocessor is capable of efficiently processing advanced vision models, for example MobileNet v2, reaching up to 400 frames per second. It has been designed in a card form factor that fits into universal M.2 slots, making it easy to add ML functionality to a variety of products, including embedded platforms and minicomputers, especially those running Linux. The module supports a wide range of hardware platforms, including those based on ARM and x86 architectures. The main features of the module include ML and AI applications, compatibility with the M.2 A+E key interface, connection via PCIe Gen2 x1, and interoperability with Linux systems (Debian 6.0+, Ubuntu 10.0+ or other derivatives). The module is powered by 3.3V with a tolerance of +/- 10%, operates at temperatures from -20 to 70°C, measures 22 x 30 x 2.3mm and weighs 3.1g. Specification:

* Application: Machine learning and artificial intelligence

* Hardware interface: M.2 A+E Key card (M.2-2230-A-E-S3)

* Interface: PCIe Gen2 x1

* Operating System: Any computer with Linux (Debian 6.0+ or derivatives - Ubuntu 10.0+)

* System architecture: x86-64 or ARM32/64 with ARMv8 instruction set

* Power supply: 3.3V DC

* Operating temperature: -20 to 70 °C

* Dimensions: 22 mm x 30 mm x 2.3 mm

* Weight: 3.1 g

* Compatibility with sockets: M.2 A+E