Description

The Coral M.2 Accelerator with Dual Edge TPU is an M.2 module (E-key) with two Edge TPU ML coprocessors, each including a PCIe Gen2 x1 interface. The Edge TPU is a small ASIC designed by Google that accelerates Tensor Flow Lite models, improving power efficiency.

Details

The Coral M.2 Accelerator is an ideal product to add powerful machine learning into an existing embedded system. Thanks to the two Edge TPUs (and thus 8 TOPS) you can double the performance of the system.

Notice: Because this module uses two PCIe x1 connections, it is incompatible with all M.2 E-key card slots. The dual Edge TPUs also result in special power requirements that you must carefully review.

Coral M.2 Accelerator Software requirements s

The Coral M.2 Accelerator with Dual Edge TPU must be operated by the Edge TPU runtime and Coral PCIe driver, which is compatible with the following systems: 

• Linux: 
  • 64-bit version of Debian 10 or Ubuntu 16.04 (or newer) 
  • x86-64 or ARMv8 system architecture 
• Windows: 
  • 64-bit version of Windows 10 
  • x86-64 system architecture 
• All systems require support for MSI-X as defined in the PCI 3.0 specification 

Why should you buy the Coral M.2 Accelerator with Dual Edge TPU?

The Coral M.2 Accelerator performs high-speed ML inferencing.

Due to the two Edge TPUs contained in this Coral M.2 Accelerator module, the inferences per second (8 TOPS) can double with various methods, such as by running two models in parallel or pipelining one model across both Edge TPUs. 

Each Edge TPU coprocessor can perform 4 trillion operations per second (4 TOPS), using 2 watts of power. For example, it can produce state-of-the-art mobile vision models such as MobileNet v2 at almost 400 FPS in a power-efficient manner.

This on-device ML processing reduces latency, increases data privacy, and removes the need for a constant internet connection. 

The Coral M.2 Accelerator is compatible with Debian Linux and Windows.

This Coral M.2 Accelerator integrates with Debian-based Linux or Windows 10 systems with a compatible card module slot.

The accelerator supports TensorFlow Lite.

This Coral M.2 Accelerator supports TensorFlow Lite models that can run on the Edge TPU. Therefore, there is no need to build models from the ground up. 

Technical Specifications

Specifications