About this item

The Coral M.2 Accelerator with Dual Edge TPU is an advanced module designed to integrate high-performance machine learning capabilities into systems with an M.2 E-key slot. This compact accelerator features two Google Edge TPU coprocessors, each capable of performing 4 trillion operations per second (4 TOPS), achieving a total peak performance of 8 TOPS. Operating at 2 TOPS per watt, it ensures power-efficient processing, making it ideal for real-time inferencing tasks. The module connects via two PCIe Gen2 x1 interfaces, allowing for parallel processing of models or pipelining, thereby enhancing throughput. It supports TensorFlow Lite, enabling the execution of optimized models directly on the device, reducing latency and dependency on cloud services. Compatible with Debian-based Linux and Windows 10 systems, the accelerator is suitable for embedded platforms, mini-PCs, and industrial gateways. Its compact dimensions (22.0 x 30.0 x 2.8 mm) and wide operating temperature range (-40°C to +85°C) make it versatile for various applications, including image and video classification, object detection, and voice recognition.