onnxruntime/docs/tutorials/iot-edge/index.md
Nat Kershaw (MSFT) 7717184965
Revert iot URL to original iot-edge (#14150)
This maintains backwards compatibility for backlinks and traffic
telemetry
2023-01-05 19:52:05 -06:00

28 lines
1.7 KiB
Markdown
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

---
title: Deploy on IoT and edge
parent: Tutorials
has_children: true
nav_order: 8
redirect_from: /docs/get-started/with-iot
---
# Deploy ML Models on IoT and Edge Devices
ONNX Runtime allows you to deploy to many IoT and Edge devices to support a variety of use cases. There are packages available to support many board architectures [included when you install ONNX Runtime](https://pypi.org/project/onnxruntime/#files). Below are some considerations when deciding if deploying on-device is right for your use case.
## Benefits and limitations to doing on-device inference
* Its faster. Thats right, you can cut inferencing time down when inferencing is done right on the client for models that are optimized to work on less powerful hardware.
* Its safer and helps with privacy. Since the data never leaves the device for inferencing, it is a safer method of doing inferencing.
* It works offline. If you lose internet connection, the model will still be able to inference.
* Its cheaper. You can reduce cloud serving costs by offloading inference to the device.
* Model size limitation. If you want to deploy on device you need to have a model that is optimized and small enough to run on the device.
* Hardware processing limitation. The model needs to be optimized to run on less powerful hardware.
## Examples
* [Raspberry Pi on Device inference](rasp-pi-cv.md)
* [Jetson Nano embedded device: Fast model inferencing](https://github.com/Azure-Samples/onnxruntime-iot-edge/blob/master/README-ONNXRUNTIME-arm64.md)
* [Intel VPU edge device with OpenVINO: Deploy small quantized model](https://github.com/Azure-Samples/onnxruntime-iot-edge/blob/master/README-ONNXRUNTIME-OpenVINO.md)