Description: Google Coral USB Accelerator - Edge TPU ML Accelerator - ARM Cortex M0.
The Coral USB Accelerator provides powerfulMLinference capabilitiesinLinux, Windows and macOS through a USB 3.0 port. The module is equipped with the ARM Cortex M0+ processor and aspecialEdge TPU Edge chip (ASIC), designedand developedby Google.Usingit, you can create, for example, modern video models such as MobileNet v2 in 100 fps, with low powerconsumption. The device cooperates with devices based on Linux distribution, such as RaspberryPi.
Coral USB Accelerator - Edge TPU ML accelerator.
Main advantages of the Google Coral USB Accelerator
- Rapid inference Edge TPU ML
- Low energy consumption
- Small dimensions
Coral is a department of the Google brand that helps to create intelligent designs based on artificial intelligence.
Main features of the accelerator
- Google Edge TPU ML accelerator coprocessor
- USB 3.0 socket type C
- Supports Linux Debian distributions as well as Windows and macOS on the host processor
- Models created using TensorFlow
- Full support for MobileNet and Inception, also possible to use external architectures
- Compatible with Google Cloud
- Local application - running ML on a TPU Edge designed by Google
Content of Coral USB accelerator kit.
Google Coral Accelerator Requirements
The device must be connected to a host computer that meets the following specifications:
- All Linux computers with a USB port:
- Debian 6.0 or later, or a derivative (e.g. Ubuntu 10.0+)
- System architecture x86_64 or ARM64 with ARMv8 kit
- Windows 10
- macOS
- Raspberry Pi
Specifications of the Google Coral USB accelerator
- Accelerator: Edge TPU ML - provides high quality ML inference for TensorFlow Lite models
- CPU: ARM Cortex M0+ 32-bit:
- Clocking: up to 32 MHz
- Flash memory with ECC: 16 kB
- RAM: 2 kB
- Connectors:
- USB 3.1 and SuperSpeed cable, 5 GB/sec transfer rate
- USB cable type C - USB A included
- Dimensions: 65 x 30 m
Useful links |
Advantages
- Powerful ML inference capabilities on Linux, Windows, and macOS
- Equipped with ARM Cortex M0+ processor and Google-designed Edge TPU chip
- Supports TensorFlow models including MobileNet v2
- High performance with low energy consumption
- Compact dimensions for easy integration into various projects
- Compatible with Raspberry Pi and other Linux-based devices
- USB 3.0 Type-C connector for fast data transfer rates (5 GB/sec)
- Full support for MobileNet and Inception architectures, as well as external architectures
- Local ML application running on Google-designed TPU Edge
- Positive user feedback on quality, reliability, and performance
Disadvantages
- Installation may be difficult for some users
AI generated based on user reviews