Coral Dev Board Mini

Coral Dev Board Mini quickly prototypes on-device machine learning (ML) products and was primarily designed as an evaluation device for the Coral Accelerator Module. The small 64mm x 48mm board is also a fully-functional embedded system that combines a single-board computer with SoC, ML, wireless connectivity, and a TPU coprocessor. The on-board TPU coprocessor can perform 4 trillion operations per second (TOPS). Each TOPS uses only 0.5 watts. It can efficiently execute state-of-the-art mobile vision models such as MobileNet v2 at almost 400 frames-per-second.

The board runs on a derivative of Debian Linux called Mendel, so many Linux tools can run on the Coral Dev Board Mini. The board supports TensorFlow Lite, eliminating the need to build models from the ground up, and supports AutoML Vision Edge that allows high-accuracy custom image classification models.

Coral, Google's platform for on-device AI, offers hardware components, software tools, and pre-compiled models for building intelligent devices.

Features:

•  - Quad-core Arm Cortex-A35 SoC CPU
•  - Google Edge TPU Coprocessor ML accelerator
•   * 4 TOPS peak performance (int8)
•   * 2 TOPS per watt
•  - ARM TrustZone Security
•  - 2 GB LPDDR3 RAM
•  - 8 GB eMMC flash memory
•  - Wi-Fi 5 (802.11a/b/g/n/ac)
•  - BLUETOOTH® 5.0
•  - USB 2.0 Type-C (2)
•  - 40-pin GPIO expansion header
•  - 64mm x 48mm
•  - Mendel Linux flashed in factory
•  - Audio
•  - 3.5mm audio headphone jack
•  - Digital PDM microphone
•  - 2.54mm 2-pin speaker terminal
•  - Video
•  - Micro HDMI (1.4)
•  - 24-pin FFC connector for MIPI-CSI2 camera (4-lane)
•  - 39-pin FFC connector for MPIP-DSI display (4-lane)

Documents:

Datasheet