Workload Consolidation with Microsoft Azure* IoT Edge for Linux* on Windows* (EFLOW)

ID 标签 723259
已更新 2/22/2022
版本 Latest
公共

author-image

作者

Microsoft Azure* IoT EFLOW Overview

Azure* IoT “Edge for Linux* on Windows*”, or simply EFLOW, enables deployment of containerized Linux workloads alongside Windows applications in Windows IoT deployments on Intel® platforms.

Users don’t need to deploy a separate Linux machine to run Linux based AI workloads as now they can run those workloads on a Windows system by virtue of EFLOW.

EFLOW is a Microsoft* technology and is delivered by Microsoft on both Windows® 10 IoT Enterprise LTSC 2019 and LTSC 2021 and will be supported by Microsoft.

EFLOW has been enhanced with Intel Integrated GPU acceleration for AI workloads using Intel® Distribution of OpenVINO™ Toolkit.  This capability is supported only on Windows 10 IoT Enterprise LTSC 2021.

For more information, please refer to Microsoft EFLOW documentation.

EFLOW Value Proposition on Intel® Platforms

EFLOW offers incredible value to developers and system integrators by combining the goodness of Microsoft Windows OS and Linux OS along with ease of deployment and maintenance of a virtualized configuration.

* Refer to EFLOW AI Inferencing with GPU Acceleration User Guide
** Refer to EFLOW Best Known Configuration (BKC) for IOTG Platforms (Microsoft Windows 10 IoT Enterprise LTSC 2021)

Intel IOTG Platforms supporting EFLOW

  • All Intel IOTG platforms that support Microsoft Windows 10 IoT Enterprise LTSC 2019:
    EFLOW + OpenVINO using CPU virtualization (Intel® Virtualization Technology) and without integrated Graphics acceleration.
  • All Intel IOTG platforms that support Microsoft Windows 10 IoT Enterprise LTSC 2021:
    EFLOW + OpenVINO using CPU virtualization (Intel® Virtualization Technology) and with integrated Graphics acceleration.

Please refer to Intel® Processor Support for Windows* for IoT Applications for full list of Intel IOTG platforms supporting Microsoft Windows Operating Systems.

Run Multiple Workloads with EFLOW on Intel Platforms

In this real-world example, a Point-Of-Sale (POS) application is running on a Windows system and the Digital Security & Surveillance (DSS) application is running on a separate Linux system. With EFLOW both the POS and DSS applications as well as the associated workloads can run on the same Intel Platform. Unleash the power and performance capabilities of Intel Platforms with accelerated AI inferencing using Intel® Distribution of OpenVINO™ Toolkit and Integrated GPU.

Industrial Use Case with EFLOW on Intel Platforms

Digital transformation in the Industrial sector is trending towards Industrial 4.0.  Automated Optical Inspection (AOI) in manufacturing is one of the Industrial use cases that heavily relies on skilled labor for visual inspection procedures in the production and assembly lines. Intel Platforms with EFLOW enable Linux based containerized Artificial Intelligent (AI) workloads to run alongside Windows based Human Machine Interface (HMI) applications.


Additional Resources