Edge AI Accelerators

Run inference on IoT end devices.

Purpose-built AI accelerators for edge inference. Run ML models on IoT devices with 4 TOPS to 200 TOPS. Ultra-low power consumption.

NPU200 TOPS< 5W

200 TOPS

Performance

< 5W

Power

TFLite/ONNX

Models

M.2/USB

Form factor

AI on every device.

200 TOPS edge inference. Under 5W.

200 TOPS

Up to 200 TOPS int8 inference. From 4 TOPS to 200 TOPS options.

Ultra-low power

Under 5W TDP. Battery-powered edge AI.

TFLite & ONNX

Run TensorFlow Lite and ONNX models natively.

M.2 & USB

M.2, USB, and PCIe form factors for any deployment.

Model compiler

Cloud compiler optimizes models for edge hardware.

OTA model updates

Update models on devices over-the-air.

Getting started

Launch your first instance in three steps. CLI, console, or API — your choice.

Terminal
ur edge ai compile my-model.tflite \
  --target=accelerator-v3 \
  --quantize=int8

Edge AI patterns.

Computer vision and voice processing.

Edge computer vision

Real-time object detection on security cameras.

View tutorial

Suggested configuration

200 TOPS · TFLite · Low power

Estimate your costs

Create detailed configurations to see exactly how much your architecture will cost. Pay for what you use, down to the second.

Configuration 1

Estimated: $44.20/mo

Edge AI Hardware

Usage Volume

K

Infrastructure

GB

Options

Premium SLA (99.99%)+25% for guaranteed availability
Config 1 cost$44.20

Cost details

$44.20

Deploy AI models to low-power edge hardware.

Configuration 1
$44.20
2× standard Replica(s)$29.20
Request Processing$10.00
Storage$5.00

Works seamlessly with

IoT Core
ML Platform
IAM
Security
Monitoring
Analytics

Frequently asked questions

AI on every device.

200 TOPS edge inference. Under 5W.