Description

The Coral Accelerator Module is designed to perform high-speed inferencing for machine learning (ML) models. The Coral module is a multi-chip module (MCM) that includes the Edge TPU ML accelerator with integrated power control, which can be connected over a PCIe Gen2 x1 or USB2 interface. The on-device ML processing reduces latency, increases data privacy, and removes the need for a constant internet connection.

Details

Key features of the Coral Accelerator Module

• Google Edge TPU ML accelerator
  • 4 TOPS peak performance (int8)
  • 2 TOPS per watt
• Integrated power management
• PCIe Gen2 x1 or USB 2.0 interface
• Surface-mounted (LGA) module
• RoHS compliant

Coral Accelerator Module functional block diagram

Why should you buy the Coral Accelerator Module?

The Coral module performs high-speed ML inferencing

The Edge TPU accelerator is a small ASIC designed by Google that accelerates TensorFlow Lite models in a power-efficient way. It is designed to perform 4 trillion operations per second (4 TOPS), using 2 watts of power—that’s 2 TOPS per watt. 

For example, one Edge TPU can execute state-of-the-art mobile vision models such as MobileNet v2 at almost 400 frames per second. 

The Coral module supports TensorFlow Lite

The Coral Accelerator Module is a surface-mounted module that includes the Edge TPU and its own power control. It provides accelerated inferencing for TensorFlow Lite models on your custom PCB hardware.

There is no need to build models from the ground up. TensorFlow Lite models can be compiled to run on the Edge TPU.

The Accelerator Module is:

• Small enough to fit into almost any design (size: 15.0 x 10.0 x 1.5 mm).

• Lets developers solder the Edge TPU acceleration onto a PCB.

• Saves even more space due to a lack of connectors or cables.

• Requires no exotic boards or surface mounts, thanks to conventional rather than high-density interconnector parts.

• Provides a standard USB 2.0 or PCIe interface to an application processor.

• Allows developers to add a single system component to their designs to enable ML acceleration.

Specifications