Google Coral Dev Board Mini


Price: $139.95
(as of Apr 18,2021 14:41:22 UTC – Details)



The Coral Dev Board Mini is a single-board laptop that lets you briefly prototype and deploy an embedded gadget with on-device ML inferencing. This board too can function an analysis machine for the Coral Accelerator Module, which lets you combine the Edge TPU into your customized PCB {hardware} as a surface-mounted module. The Edge TPU is a small ASIC designed via Google that speeds up TensorFlow Lite fashions in an influence environment friendly approach: each and every one is able to appearing 4 trillion operations in line with 2nd (4 TOPS), the use of 2 watts of energy—that is 2 TOPS in line with watt. For instance, one Edge TPU can execute state of the art cell imaginative and prescient fashions akin to MobileNet v2 at virtually 400 frames in line with 2nd. This on-device ML processing reduces latency, will increase information privateness, and eliminates the will for a relentless web connection. Technical specs: MediaTek 8167s SoC (Quad-core Arm Cortex-A35), IMG PowerVR GE8300 built-in GPU, ARM TrustZone Security, Wi-Fi 5, Bluetooth 5.0, 8 GB eMMC, 2 GB LPDDR3, USB 2.0 Type-C OTG, HDMI 1.4a (micro), MIPI DSI show, MIPI CSI digicam, 3.5 mm headphone jack, Digital PDM microphone, 2.54mm 2-pin mono speaker terminal, 40-pin GPIO enlargement header, Mendel Linux flashed in manufacturing unit.

The Coral Dev Board Mini is a single-board laptop that lets you briefly prototype and deploy an embedded gadget with on-device ML inferencing.
The board comprises the Edge TPU coprocessor, which is a small ASIC designed via Google that speeds up TensorFlow Lite fashions in an influence environment friendly approach. It’s able to appearing 4 trillion operations (tera-operations) in line with 2nd (tops), the use of 0.5 watts for each and every tops (2 tops in line with watt).
Provides a whole gadget: a single-board laptop with SoC + ML + wi-fi connectivity, all at the board operating a by-product of Debian Linux we name Mendel, so you’ll run your favourite Linux gear with this board.
Supports TensorFlow Lite: no wish to construct fashions from the bottom up. Tensorflow Lite fashions will also be compiled to run at the Edge TPU.
Supports AutoML Vision Edge: simply construct and deploy rapid, high-accuracy customized symbol classification fashions in your machine.
MediaTek 8167s SoC (Quad-core Arm Cortex-A35)
2 GB LPDDR3 and eight GB eMMC reminiscence