Special B2B pricing available after sign-in
Sign in Register
0Cart

Google Coral USB Accelerator Edge TPU, 4 TOPS, USB 3.0, TensorFlow Lite

Brand:
Raspberry
Google Coral USB Accelerator für Raspberry Pi | GOOGLE CORAL USB
Google Coral USB Accelerator Edge TPU, 4 TOPS, USB 3.0, TensorFlow Lite
€152.59 inc. VAT
Buy
GOOGLE CORAL USB
€128.23
€152.59 inc. VAT
Ships from EU stock: 18 item(s)
+ −
Add to wish list Ask a question
Product Code
CF40122
EAN Code
0842776110077
Manufacturer
Raspberry

Quick Specs

Supported model format
TensorFlow Lite models
USB transfer rate
SuperSpeed, 5 Gbps
Host interface
USB 3.0 / USB 3.1 Gen 1
Required host port
One available USB port
Delivery

Order Today...
Ships...
Delivery...
ℹ️ Large or fragile items (e.g. monitors) may require extra handling time. We'll keep you informed.
Payment options
✓ Bank Deposit
Secure checkout with flexible payment methods
Trusted Tech Choice
Buying Benefits Verified
Competitive EU pricing
Delivered to Cyprus from EU stock
Local warranty support
Singular Selected

Google Coral USB Accelerator Edge TPU - 4 TOPS on-device AI

The Google Coral USB Accelerator adds Edge TPU machine-learning inference to Raspberry Pi and other compatible computers through USB. With up to 4 TOPS of TensorFlow Lite inference performance and 2 TOPS per watt, it is designed for developers building or running supported on-device AI applications.
TensorFlow Lite models SuperSpeed, 5 Gbps USB 3.0 / USB 3.1 Gen 1 One available USB port

The Google Coral USB Accelerator adds Edge TPU machine-learning inference to Raspberry Pi and other compatible computers through USB. With up to 4 TOPS of TensorFlow Lite inference performance and 2 TOPS per watt, it is designed for developers building or running supported on-device AI applications.

Best For
✓ A good fit for developers
✓ Armv7, or Armv8 host

Key Features

 
Commercial Fit
The main reason to choose this accelerator is its dedicated Google Edge TPU, providing up to 4 TOPS of supported TensorFlow Lite inference through a compact USB connection without replacing the host computer.
 
AI Inference Performance
AI Inference Performance: 4 TOPS
 
Cable Length
300 mm (12 in) | ±20 mm tolerance | 300 mm (12 in), ±20 mm tolerance
 
EAN / UPC
EAN / UPC: 0842776110077

Technical Specifications

Product Type USB machine-learning accelerator
ML Accelerator Google Edge TPU ASIC
AI Inference Performance 4 TOPS
Energy Efficiency 2 TOPS per watt
Supported Model Format TensorFlow Lite models
Host Interface USB 3.0 / USB 3.1 Gen 1
USB Transfer Rate SuperSpeed, 5 Gbps
Required Host Port One available USB port | USB 3.0 recommended for best performance
Included Cable USB Type-C to USB Type-A
Cable Length 300 mm (12 in) | ±20 mm tolerance | 300 mm (12 in), ±20 mm tolerance
Supported Operating Systems Linux Debian 10 or derivative such as Ubuntu 18.04 | macOS 10.15 with MacPorts or Homebrew | Windows 10
Supported Host Architectures x86-64 | Armv7 32-bit | Armv8 64-bit | x86-64, Armv7 32-bit, or Armv8 64-bit
Raspberry Pi Support Supported | Google tested Raspberry Pi 3 Model B+ and Raspberry Pi 4
Python Requirement Python 3.5, 3.6, or 3.7
Manufacturer Google
Model / Part Number G950-01456-01
EAN / UPC 0842776110077
SKU / EAN G950-01456-01 | 0842776110077

Frequently Asked Questions

What should I know before buying?
This is an accelerator, not a standalone computer, so a compatible host and an available USB port are required. USB 3.0 is recommended for best performance, and the supported model format is TensorFlow Lite. Check that your operating system and Python version match the listed requirements.
What connectivity does it offer?
USB 3.0 / USB 3.1 Gen 1
What is it compatible with?
TensorFlow Lite models
Singular Expert Insight
The main reason to choose this accelerator is its dedicated Google Edge TPU, providing up to 4 TOPS of supported TensorFlow Lite inference through a compact USB connection without replacing the host computer.

Buyer’s Guide

Buy it if
A good fit for developers and makers who need on-device TensorFlow Lite inference on a Raspberry Pi or another compatible x86-64, Armv7, or Armv8 host.
Skip it if
This is an accelerator, not a standalone computer, so a compatible host and an available USB port are required. USB 3.0 is recommended for best performance, and the supported model format is TensorFlow Lite. Check that your operating system and Python version match the listed requirements.

Features - Raspberry - GOOGLE CORAL USB

More Details
Product Code
CF40122
EAN Code
0842776110077
Manufacturer
Raspberry
Find similar
  • Recently Viewed
  • Customers Also Bought
 
Why Buy from Singular

Trusted by businesses across Cyprus

Genuine products, EU stock, and expert support for every purchase.
✓ Genuine Products
Sourced through official distribution channels, with full manufacturer warranty.
✓ EU Stock
Shipped from our EU distribution network, with real-time availability on every product.
✓ B2B Ready
VAT invoices for business deduction and credit terms for verified buyers.
Expert Support
Talk to our team for sizing, configuration or volume orders.